University of Southern California Dissertations and Theses

A variationist study of relative clauses in Spanish

(USC Thesis Other)

A variationist study of relative clauses in Spanish

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION TO USERS
This manuscript has been reproduced from the microfilm m aster. UMI fiims
the text directly from the original or copy submitted. Thus, som e thesis and
dissertation copies are in typewriter face, while others may be from any type of
computer printer.
The quality o f this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality illustrations
and photographs, print bleedthrough, substandard margins, and improper
alignment can adversely affect reproduction..
In the unlikely event that the author did not send UMI a com plete m anuscript
and there are missing pages, th ese will be noted. Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand com er and continuing
from left to right in equal sections with small overlaps.
Photographs included in the original manuscript have been reproduced
xerographically in this copy. Higher quality 6” x 9” black and white
photographic prints are available for any photographs or illustrations appearing
in this copy for an additional charge. Contact UMI directly to order.
ProQuest Information and Learning
300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA
800-521-0600
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
A VARIATIONIST STUDY OF RELATIVE
CLAUSES IN SPANISH
by
Kareen Liliana Gervasi
A dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
in Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(SPANISH)
MAY 2000
Copyright 2000 Kareen Liliana Gervasi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 3018002
___ ®
UMI
UMI Microform 3018002
Copyright 2001 by Bell & Howell Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
Bell & Howell Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UNIVERSITY OF SOUTHERN CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES, CALIFORNIA 9 0 0 0 7
This dissertation, written by
under the direction of h.£T....... Dissertation
Committee, and approved by all its members,
has been presented to and accepted by The
Graduate School in partial fulfillment of re
quirements for the degree of
DOCTOR OF PHILOSOPHY
Kareen Liliana Gervasi
Dean of Graduate Studies
DISSERTATION COMMITTEE
5 *
Chairperson
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ii
ACKNOW LEDGEM ENTS
I would like to thank all the people who helped me during my graduate studies at
the Department of Spanish and Portuguese, particularly during the process doing the
research and writing of this dissertation.
I wish to express my most sincere gratitude to Dr. Carmen Silva-Corvalan for her
invaluable help and advice during my graduate studies and throughout all the stages of the
completion of this dissertation.
I want to thank Professor Mario Saltarelli and Professor Edward Finegan for their
opportune suggestions and ideas during the research and writing of this work.
I also would like to express my gratitude to Dr. Gayle Vienna for her support and
advice regarding my teaching in the Spanish Department.
I wish to thank my friend Deborah Gill for her ideas, her constant support, and
most o f all, her friendship.
Finally, I would like to express my thanks to Teeanna Rizkallah for her help in
editing this dissertation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table of Contents
Pages
Acknowledgements ii
List of tables v
Chapter 1: Introduction I
1.1. Problem and objectives 1
1.2. Theoretical background 4
1.2.1. Types o f relative clauses and types
o f relativizers in Spanish 4
1.2.2. Syntactic functions of relative
clauses and their antecedents 16
1.2.3. The Accessibility Hierarchy for relative
clause formation: the case of Spanish 26
1.3. Linguistic variables 3 0
1.4. Social variables 33
1.5. Situational Variables 38
1.5.1. Register 38
1.6. Summary 39
Chapter 2: Methodology 41
2.1. The speakers and the data 41
2.2. Coding for analysis o f linguistic factors 43
2.2.1. Restrictive/Nonrestrictive
relative clause 46
2.2.2. Humanness 49
2.2.3. Definiteness 49
2.2.4. Information Status 50
2.2.5. Grounding 57
2.2.6. Functional Role of
the relative clause 64
2.2.7. Informational value of
the relative clause 69
2.3. Presence/Absence o f a Preposition 71
2.4. Register 74
2.5. Social variables 75
Chapter 3: Results of the analysis of Linguistic variables 77
3.1. Introduction 77
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
iv
3.2. Other studies 78
3.3. Analysis 78
3.3.1. Humanness 78
3.3.2. Grounding and Humanness 85
3.3.3. Functional role o f the relative clause 111
3.4. Conclusions 156
Chapter 4: Discourse and Social variables 166
4.1. Introduction 166
4.2. Restrictive and Nonrestrictive relative clause,
Level of education, and Socioeconomic level 169
4.3. Presence/Absence o f a preposition, Level of
education, Socioeconomic level, and Register 173
4.4. Presence/Absence of a preposition and
Monolingual/Bilingual 180
4.5. Summary and conclusions 183
Chapter 5: Conclusions 186
5.1. Semantic/Pragmatic factors 186
5.1.1. Humanness 186
5.1.2. Humanness, Grounding, and Type of
Combination 188
5.1.3. Functional Role o f the RC and Type of
Combination 191
5.2. Social and Discourse variables 196
5.2.1. Restrictive and Nonrestrictive RCs and their
correlation with social factors 196
5.2.2. Presence/Absence o f a preposition
with social variables and Register 198
5.2.3. Presence/Absence o f a preposition
and Bilingual/Monolingual speakers 199
5.3. Final remarks 199
Bibliography 202
Appendix 1: The different TCs syntactic roles of the
RCs and their antecedents 205
Appendix 2: Types of registers in this study 214
Appendix 3: Social variables in this study 215
Appendix 4: Humanness o f the element that provides Grounding
(Nonhuman and Human antecedents) 216
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
V
LIST OF TABLES
CHAPTER 2
Table 1 Functional roles of a Relative Clause, Information Status o f an antecedent,
and Informational Value of a Relative Clause. 71
CHAPTER 3
Table 2 Type of Combination by Humanness. 80
Table 3 Grounding by Type of Combination by Human antecedents. 88
Table 4 Grounding by Type of Combination o f Nonhuman antecedents. 90
Table 5 Functional Role o f the Relative Clause by Type of Combination for
Human antecedents. 115
Table 6 Functional Role of the Relative Clause by Type of Combination for Human
antecedents. 117
Table 7 Functional Role of the Relative Clause by Type of Relative Clause by
Anchoring (Human and Nonhuman antecedents). 124
Table 8 Functional Role of the Relative Clause by Type o f Relative Clause by
Anchoring (Nonhuman antecedents). 129
Table 9 Functional Role of the Relative Clause by Type of Combination with
Main-clause grounding (Human and Nonhuman antecedents). 131
Table 10 Functional Role o f the Relative Clause by Type of Relative Clause with
Main-clause grounding (Human antecedents). 134
Table 11 Functional Role of the Relative Clause by Type of Relative Clause with
Main-clause grounding (Nonhuman antecedents). 136
Table 12 Functional Role of the Relative Clause by Type of Combination by
Proposition-linking grounding (Human and Nonhuman antecedents). 138
Table 13 Functional Role o f the Relative Clause by Type of Relative Clause with
Proposition-linking grounding (Human and Nonhuman antecedents). 142
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
VI
Table 14 Functional Role o f the Relative Clause by Type of Combination by
Proposition-linking grounding (Human antecedents).
Table 15 Functional Role o f the Relative Clause by Type of Combination by
Proposition-linking grounding (Nonhuman antecedents).
Table 16 Functional Role o f the Relative Clause by Type of Relative Clause (Human
antecedents).
Table 17 Functional Role o f the Relative Clause by Type of Relative Clause
(Nonhuman antecedents).
Table 18 Functional Role o f the Relative Clause by Type of Relative Clause by
Proposition-linking grounding (Nonhuman antecedents).
CHAPTER 4
Table 19 Frequency of production of Restrictive and Nonrestrictive Relative
Clauses among monolingual speakers with College and No College
education.
Table 20 Frequency of production of Restrictive and Nonrestrictive Relative
Clauses among monolingual speakers from the middle and lower class.
Table 21 Presence/Absence of a Preposition with the relativizer "que/that" by Level
of Education (Monolingual speakers).
Table 22 Presence/Absence of a Preposition with the relativizer "donde/where" by
Level of Education (Monolingual speakers).
Table 23 Presence/Absence of a Preposition with the relativizer "que/that" by
Socioeconomic Level (Monolingual speakers).
Table 24 Presence/Absence of a Preposition with the relativizer "donde/where" by
Socioeconomic Level (monolingual speakers).
Table 25 Presence/Absence of a Preposition with the relativizer "que/that" by Type
of Register among Middle Class speakers.
Table 26 Presence/Absence of a Preposition with the relativizer "donde/where" by
Type of Register among Middle Class speakers.
144
148
158
159
162
171
172
174
175
176
177
178
178
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 27 Presence/Absence of a Presposition with the relativizer "que/that" by
Register among College educated speakers.
Table 28 Presence/Absence of a Preposition with the relativizer "que/that" among
Monolinguals with No college education, Monolinguals with College
education, and Bilinguals.
Table 29 Presence/Absence of a Preposition with the relativizer "donde/where"
among Monolinguals with No college education, Monolinguals with College
education, and Bilinguals.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
viii
Kareen Gervasi Carmen Silva-Corvalan
A VARIATIONIST STUDY OF RELATIVE CLAUSES IN SPANISH
This dissertation examines correlations between the grammar of Spanish relative
clauses, semantic/pragmatic factors such as humanness, definiteness, information status of
noun phrases, the functional role of relative clauses, and social factors. I follow in part
Fox & Thompson’s (1990) study of the grammar o f relative clauses in English, but include
a wider range of syntactic possibilities involved in relativization in Spanish. The
correlation with social variables is examined with respect to the use of restrictive and
nonrestrictive relative clauses and the variable use o f prepositions with a relativizer. The
use of a preposition with a relativizer is analyzed across various types of registers in the
speech of 98 individuals from Mexico City and from the southwestern United States. The
data are analyzed by submitting them to some of the programs in the Statistical Package
for the Social Sciences (SPSS). The results highlight the importance of a variationist
approach that includes semantic/pragmatic, discourse and social factors in the
investigation o f grammatical variation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C H A PTER 1
1
INTRODUCTION
1.1. Problem and objectives
This study searches for syntactic and functional explanations of grammatical
patterns in Restrictive and Nonrestrictive relative clauses (RCs) in Spanish in relation to
semantic/pragmatic factors. It also examines the distribution of RCs across various social
and education groups. Fox and Thompson define information flow as "the interactionally
determined choices that speakers make which determine intonational, grammatical, and
lexical choices." (299) One o f the aspects o f information flow that Fox and Thompson
considered is Grounding, which refers to the ways speakers make noun phrases relevant
in the discourse. Other important factors in their study are Information Status of the noun
phrase antecedent of the RC, Definiteness of the noun phrase antecedent of the RC, and
the Function of the RC with respect to its antecedent. The information flow factors that
will be considered in this work are Information Status of noun phrases, Grounding,
Humanness, Definiteness, and Function of the RC.
In data from English conversation, Fox and Thompson examined the attention paid
by the speakers to information flow factors and how such attention determines structural
choices in RC constructions. I will examine the influence of these factors in the
production of RCs in Spanish.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2
The three main objectives o f this dissertation are:
(a) To explain the information flow factors that underlie the distribution of Restrictive
and Nonrestrictive RCs. The research done by Fox and Thompson in conversational
English provides an explanation of how the choices speakers make in the structure of
relative clauses reveal the existence of factors involved in the grammatical choices these
speakers make in the process o f relativization. These are Information Status of nominal
phrases, Grounding, Definiteness, Humanness, and Function o f the RC.
For the purposes of their study, Fox and Thompson isolated two possible
functional types of RCs in their corpus (301-2). In the first kind, the RC gives a
characterization or description of a New noun phrase antecedent (one that the hearer did
not know previously). The second type o f RC "helps to identify a Given Head NP
referent, previously known to the hearer." (302) The examples provided by the authors1
are:
Characterization
(1) There’s a woman in my class [who’s a nurse].
Identification
(2) and then the one [that’s bigoted], she’s married to this guy
‘ The noun phrase antecedents of the relative clauses in the present study
are underlined, and the relative clauses will be placed within brackets ('[]'). This
convention will also be followed when relative clauses are cited from other authors
or sources.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3
I will investigate what role the aforementioned semantic/pragmatic factors play in
the structure of relative clauses in my data.
(b) To examine the speakers' use of Restrictive and Nonrestrictive RCs and the
variable use of prepositions with a relativizer in two levels of education and in two
socioeconomic levels.
Macaulay’s investigation in Ayr, Scotland, revealed, for instance, that WH relative
markers (e.g. which, who) were very common among middle class speakers but rare
among lower-class speakers (64-65). There is also a difference between these two groups
o f speakers in terms o f the deletion o f subject relative pronouns in Restrictive RCs ( ex:
"and there was a select number o f us $ became members.") (68) This speech pattern
occurs in less than 3 percent (5/180) of the middle class, but in 24 percent (53/221) of the
lower class) (Macaulay 68).
Macaulay found that middle class speakers tend to use Nonrestrictive RCs more
frequently than the lower class speakers (20% o f Nonrestrictive RCs among the middle
class speakers (84/423), and 5% o f Nonrestrictive RCs among the lower class speakers
(24/526)) (Macaulay 64).
In the present study, I analyze two socioeconomic levels, middle and lower, among
speakers in both Mexico City. I also study the speech of speakers from the southwestern
United States.2
T he speakers from the southwestern United States are from San Marcos,
TX; Mora, NM; Tucson, AZ; and San Jose, CA.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(c) To study the presence/absence of a preposition with a relativizer in different types
o f communicative situations or registers in which the speakers in the present study
participate.
The use o f a preposition in contexts in which its use is variable will be considered
standard variant, rather than the absence of a preposition in the same contexts. I analyze
the influence that different communicative situations may have on the use of the standard
variant.
1.2. Theoretical background.
1.2.1. Types of relative clauses and types of relativizers in Spanish
In order to provide a definition of an RC, I will refer to two normative Spanish
grammars: Real Academia Espanola and Bello. The definition o f RC in Comrie, given
from a typological point of view, will also be considered in this work.
Real Academia Espanola locates the RC within adjectival subordination.
According to Real Academia Espanola: "a noun, whichever is the syntactic function that
it plays, can take a sentential complement introduced by a relative pronoun." (5243 ) One
of the examples given is (3):
(3) La distancia es oportuna acaso para amortiguar el golpe [que te doy...]/llorando.
(A. Nervo, Cuentos Misteriosos: El heroe).
'The distance is opportune perhaps to soften the punch [I give to you ...j/crying.1
M y translation of Real Academia Espanola (524).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5
In (3), the noun phrase "el golpe/the punch" has an RC, which is introduced by the
relative pronoun "que" (Real Academia Espanola 524).
The RCs that specify and restrict an antecedent (Restrictives) and the RCs that
only point at a quality or circumstance of such antecedent (Nonrestrictives) are illustrated
in (4) and (5), respectively:
(4) Todas las casas [que hemos visto] son pequenas (Real Academia Espanola 525)
All of the houses [that we have seen] are small
(5) Juan, [que es diligente], trabaja sin descanso (Real Academia Espanola 525)
Juan, [who is responsible], works without a break
From a typological point of view rather than from a normative one, Comrie also
distinguishes between Restrictives and Nonrestrictives. He considers that "restrictive
relative clauses are more central to the notion of relative clause than are nonrestrictives..."
(Comrie 143) and provides a clear semantic definition of a restrictive RC: "A relative
clause then consists necessarily of a head and a restricting clause. The head in itself has a
certain potential range of referents, but the restricting clause restricts this set by giving a
proposition that must be true of the actual referents, of the over-all construction." (Comrie
143)
Other criteria that both Bello and Real Academia Espanola mention in order to
distinguish between Restrictive and Nonrestrictive RCs are that while the latter can be
omitted without changing the truth o f the main clause, Restrictives cannot. Furthermore,
in the Restrictives, both antecedent and RC are closely linked in the enunciation, whereas
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
in the Nonrestrictives or explicatives they are separated by a pause, which is represented in
written texts by a comma (Real Academia Espanola 525-26).
As Bello points out, in "recitation" many times the restrictive and nonrestrictive
meaning are only distinguished by the pause that is made in the Nonrestrictive (Bello 300),
as (6) and (7) illustrate:
(6) Las senoras. [que deseaban descansar], se retiraron
The ladies, [who wanted to rest], left.
In (6), the RC is a Nonrestrictive, but if the pause in the conversation and the
comma in writing are suppressed, the example becomes a Restrictive one, as in (7):
(7) Las senoras [que deseaban descansar], se retiraron
The ladies [that wanted to rest] left
In (6), the set of referents o f the antecedent "las seiioras/the ladies" is wider than
the set of referents of the construction in (7), "las senoras que deseaban descansar/ the
ladies who wanted to rest." In (7), the RC is necessary to identify a smaller set of referents
than the one in (6), so it can be understood that we refer to the ladies "who wanted to
rest" only, and not to a wider group.
In an observation about the functions of Restrictives and Nonrestrictives, Campos
notes that the information contained in a Restrictive is "crucial" for the identification of its
antecedent (98). On the other hand, a Nonrestrictive RC provides additional information
about an antecedent that has already been identified. In the present study, I prefer to
define a Restrictive RC as being crucial for the specification o f the referent of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7
antecedent. I use the term "identification" to refer to one o f the various functional roles
that an RC can have with respect to its antecedent.
The following is a summary o f the differences between Restrictive and
Nonrestrictive RCs that are noted in the literature. The purpose of the summary is to
recall the main criteria proposed to distinguish between Restrictives and Nonrestrictives:
Restrictives
(a) They specify and restrict a set of referents in an antecedent (Real Academia Espanola;
Bello; Comrie; Campos).
(b) Information provided by them is "crucial" for the specification of an antecedent
("identification" in Campos (98)), as illustrated in (84 ):
(8) Las personas [que conozco] me Ilamaron
'The persons [that I know] called me up'
The presence o f the RC "[que conozco/that I know]" is crucial to specify the set of
individuals who called, which consists o f only "the persons that I know" and not "all the
persons."
In (9), because o f the use of a proper noun, a Restrictive RC is not needed. The
use o f a Restrictive RC makes the construction unacceptable, as (9) and (10) (my
examples) illustrate:
“ The examples in the present investigation that do not contain a citation are
mine.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(9) Ana [que usa mucho maquillaje] es la hija mayor
'Ana [who wears a lot of makeup] is the oldest daughter.'
The same can be said in (10) about pronouns as antecedents:
(10) Ella [que usa mucho maquillaje] es la hija mayor
'She [who wears a lot of makeup] is the oldest daughter.'
(c) As Campos notes, Restrictives allow a syntactically non-expressed antecedent
(99). Campos' example in (11) illustrates this:
(11) Los (f) [que estudian trances] iran a Francia (Campos 99)
'The ones [who study French] will go to France.'
The article "los" (the/plural) agrees, in gender and number, with a zero antecedent,
which is understood by the context.
(d) More than one Restrictive RC can modify an antecedent (Campos 99). For
English, McCawley also notes that restrictives "can be 'stacked.’" (419) Campos'
example is shown in (12):
(12) Solo los muchachos [que estudian trances] [que sacan buenas notas] podran
participar en el viaje a Francia (Campos 99)
'Only the guvs [who study French] [who get good grades] will be able to participate
in the trip to France.'
Nonrestrictives
(a) Provide additional information about an antecedent (Real Academia Espanola;
Bello; Campos).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
9
(b) Information provided is "additional" information about an "already specified
antecedent" ("identified" in Campos (98)), as in (13):
(13) Miguel, [que es arquitecto], trabaja en Miami
MigueL [who is an architect], works in Miami'
A proper noun like "Miguel," with a unique referent, can only be an antecedent o f
a Nonrestrictive. If the Nonrestrictive RC is deleted, the message conveyed by the main
clause will be the same, that is, that a person called Miguel works in Miami.
(c) Nonrestrictives do not allow a syntactically non-expressed antecedent (Campos
99). If (11), noted above, becomes nonrestrictive, the result will be unacceptable,
as illustrated in (14):
(14) Los (f>, [que estudian ffances], iran a Francia
'The ones, [who study French], will go to France'
In (14) without a Restrictive RC, we are adding extra information about an
antecedent that has not been specified.
(d) Campos proposes that only one Nonrestrictive RC can modify an antecedent (99),
just as McCawley notes that Nonrestrictives cannot be "stacked." (419) Campos' example
is shown in (15). Campos considers example (15) ungrammatical. However, (15) is
perfectly understandable as a series o f comments made about an already identified
antecedent.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10
(15) * Marcos, [que estudia frances], [que saca buenas not as], participara en el viaje a
Francia (Campos 99)
"Marcos, [who studies French], [who gets good grades], will participate in the trip
to France'
Normative grammars (Real Academia Espanola; Bello; Alarcos) consider "que,"
"cual," "quien," and "cuyo" as relativizers in Spanish. The relative pronoun "que,"
according to Real Academia Espanola, "is invariable for gender and number, and can refer
to one or several antecedents of person or thing...." (528) Although this relative pronoun
can be substituted by "el cual" in the Nonrestrictives, this cannot be done in the
Restrictives (Real Academia Espanola 529). Real Academia Espanola says that "cual" is
used as a relative equivalent to "que/that" when it is preceded by the article, forming the
groups "el cual," "la cual," "lo cual," "los cuales," "las cuales," and in general terms, in
nonrestrictive sentences (my translation) (530), as in (16) and (17), from Real Academia
Espanola:
(16) Los estudiantes, [que estaban Iejos], no veian la pizarra (Real Academia Espanola
529)
'The students, [who were far away], could not see the board'
(17) Los estudiantes. los [cuales estaban lejos], no veian la pizarra (Real Academia
Espanola 529)
'The students, [who were far away], could not see the board'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
The type of relativizer that is clearly predominant in my investigation is "que/that,
which, who, whom" (3094 cases out o f3345 Restrictive and Nonrestrictive RCs5 ). The
other types found in my data were "donde/where" (219 cases), "cual/which, who" (14
cases), "quien/who" (11 cases) "cuando/when" (6 cases), and "cuyo/whose" (1 case).
Real Academia Espanola also refers to the mandatory use o f prepositions when the
relative "que" is a circumstantial complement (529s). According to Real Academia
Espanola, when the relativizer "que" has the function of circumstantial complement, it has
to be accompanied by the preposition that is appropriate for the type of complement, as
illustrated in (18), from Real Academia Espanola (529):
(18) Se pasaron quince dias [en que no la vimos] (Cervantes, Quijote, I, 40)
Tifteen days passed [in which we did not see her]'
However, this norm is often violated by literary texts and colloquial speech (Real
Academia Espanola 529).
Real Academia Espanola also refers to the omission of a preposition when it is the
same as the preposition that is contained in the antecedent. This is shown in (19), from
Real Academia Espanola:
5 This includes human, nonhuman, and animal antecedents.
Circumstantial complements indicate "place", "time", "mode", "cause",
"means", or "instrument" o f the action expressed by the verb (Real Academia
Espanola 375).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 2
(19) Nos vamos vestidos con los mismos vestidos [(j) que representamos] (Real
Academia Espanola 529)
W e go out dressed up with the same clothes [that we wear in our performances]'
The preposition omitted in (19) is "con/with."
The omission of a preposition is also frequent when circumstances of time and
place are expressed in the antecedent (Real Academia Espanola 529). In (20), from Real
Academia Espanola, "en/in" has been omitted in the RC, since a circumstance of place is
being expressed in the antecedent "el Iugar."
(20) En el lugar [que fire fundada Roma], no se veran... (Real Academia Espanola 529)
'In the place [in which Rome was founded], they could not be seen...'
Another use not accepted by Real Academia Espanola is the omission o f the
preposition "a" of the dative or the accusative of person in cases in which the relative
pronoun "que" is accompanied by an unstressed pronoun that has the same grammatical
case (529-30). This is shown in (21), from Real Academia Espanola, in which the
preposition "a" is omitted, and "Ie" is the unstressed pronoun with an indirect object
function:
(21) Uno [que le llaman el chato] (Real Academia Espanola 530)
'One [whom they call "shorty"]1
In the present investigation, cases of presence and absence of a preposition were
found with the relativizers "que" and "donde."
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
The relativizer "quien" (Real Academia Espanola 531-32) "is equivalent to "el
que," "la que," "los que," "las que." It refers to persons only or personified things,7 as can
be seen in (22):
(22) Las personas [a quienes invitamos] llegaron muy tarde
'The persons [that we invited] arrived very late'
As I noted earlier, the data in this investigation reveal that "quien" has a very
limited use. This may be due to the fact that the relativizer "que" can be used both with
Human an Nonhuman antecedents.
As for "cuyo," it has different characteristics with regard to its agreement (in
gender and number) with a noun. Real Academia Espanola characterizes this relativizer in
the following way: "It is the possessive adjective o f the relative pronouns, like "mio"
(mine), "tuyo" (yours), etc., are possessive personal pronouns, and it is equivalent because
of its meaning, to "de quien", "del cual", "de lo cual"...." (532)
This relative pronoun joins two nouns, one o f them the main clause (antecedent)
and the other one the subordinate clause, and always indicates the person or thing that is
possessed or a characteristic of the antecedent (Real Academia Espanola 532-33).
Example (23) illustrates this:
7 A 1 1 translations are mine unless noted otherwise.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
(23) La muchacha [cuya madre vino] es doctora
'The girl [whose mother came over] is a doctor1
The data in this investigation revealed that on six occasions, the speakers used the
construction "que+su" instead o f the relativizer "cuyo," as is illustrated in (24):
(24) ...Por alia vive una sefiora [que hace muchos afios, a su papa lo metieron a la
carcel...] (Lope Blanch, El Habla Popular 188).
'...A woman [whose father was jailed many years ago...] lives over there...1
In (24), the speaker could have said: "...una senora [cuyo papa Io metieron hace
muchos afios a la carcel...]."
It is necessary to add to the list o f relative pronouns those that Real Academia
Espafiola, Bello, and Alarcos Llorach call relative adverbs. This group consists of
"donde" (where), "cuando"' (when), "como"' (how), and "cuanto"8 (as much as, as many
as, all that).
The relative adverb o f place, "donde," is used with antecedents which indicate
place (Real Academia Espafiola 533), as (25), from Real Academia Espafiola illustrates:
(25) La casa [donde pase mi nifiez] no existe ya (Real Academia Espafiola 533)
'The house [where I spent my childhood] does not exist anymore'
Real Academia Espafiola's example o f the use of "cuando" (relative adverb of
time) as a relative adverb is shown in (26):
8 Alarcos Llorach, unlike Real Academia Espafiola and Bello, does not
include "cuanto" in the group o f relative adverbs. Rather, he considers "cuanto" "a
quantifier of generalization" (Alarcos Llorach 103).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
15
(26) Recordabamos los afios [cuando ibamos juntos a la escuela] (Real Academia
Espafiola 534)
W e would remember the years [when we would go to school together]'
The relative adverb "como" has, according to Real Academia Espafiola, a modal
meaning, and it is used with antecedents that express mode or manner, means, procedure,
etc. (534). The example given by Real Academia Espafiola is (27):
(27) Estaban de acuerdo sobre la manera [como habia de entablarse la demanda] (Real
Academia Espafiola 534)
'They agreed on the manner [in which the demand was supposed to be done]'
Real Academia Espafiola calls "cuanto" a relative of generalization (534). In (28),
from Real Academia Espafiola, the relative adverb "cuanto" (all that) has "todo"
(everything) as its antecedent.
(28) Todo [cuanto decia] le parecia gracioso (Real Academia Espafiola 534)
'All [that they would say], he found funny.'
In the following section, the types of RCs and types of relativizers will be used to
establish a classification of Spanish RCs according to the syntactic functions that both
Restrictive and Nonrestrictive RCs and their antecedents can have. Having a clear idea of
the possible types of RCs in Spanish is crucial to systematically analyze the different
examples in the data.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
1.2.2. Syntactic functions of RCs and their antecedents
In this section I will discuss the different syntactic functions that RCs can have in
Spanish. When normative grammars refer to syntactic functions regarding RCs, they
mention some of the functions that relative pronouns can have. Real Academia Espafiola
notes that relative pronouns can have different functions within the RC (525). In (29),
Real Academia Espafiola (525) shows this fact:
(29) Veo aqul — dijo — un hombre rubio. [que no le quiere a usted] (A. Nervo,
Cuentos)
'I see here — he/she said — a blond man, [who doesn' t like you]1
The antecedent of the RC, "un hombre rubio/a blond man," has the role of direct
object of the verb "veo/I see." On the other hand, within the RC introduced by the relative
pronoun "que," this relative pronoun has the syntactic function of subject.
The relative pronouns can have, within their RCs, syntactic functions that are
characteristic of nouns, adjectives, and adverbs (Alarcos Llorach 94). In what follows, I
illustrate the distribution of "quien," "que," "el cual," "cuyo," "donde," and "cuando."
(a) Quien
The relative "quien"1 (who) can only function as a noun, and varies according to
the number, "quien" for singular and "quienes" for plural. It is used with Human
antecedents or with Nonhuman antecedents that are referred to as Human9 in the discourse
9 As in the example:
No era la fortaleza [...] [quien la sacaba del desierto] (Alarcos Llorach 99)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
17
(Alarcos Llorach 99). "Quien" can only have, within its RC, the syntactic functions that
are characteristic of nouns, that is, "Subject," "Direct Object," "Indirect Object," "Object
o f Preposition," and "Circumstantial." (Alarcos Llorach 99)
In what follows, the possible functions for the relativizer "quien" are shown for
Restrictives and Nonrestrictives:
The relative "quien" can be used with the syntactic functions illustrated in what
follows:
Restrictive RCs
(30) El muchacho [quien viene ahi] es Pedro (Subject)
'The guv [who is coming] is Pedro1
(31) La persona [a quien visitamos ayer] es enfermera (Direct Object)
'The person [whom we visited yesterday] is a nurse'
(32) El estudiante [a quien le dieron el primer premio] estaba muy contento (Indirect
Object)
'The student [to whom they gave the first prize] was very happy'
(33) La chica [de quien hablamos ayer] viene hoy (Object of Preposition)
’ The girl [whom we were talking about yesterday] is coming today'
(34) La personas [por quienes crearon esos programas de salud] son menores de edad
(Circumstantial Complement of "cause")
'The persons [for whom they created those health programs] are minors'
'It was not the strength [...] [that was taking her out of the desert...]'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
Nonrestrictive RCs
(35) Andrea, [quien vive cerca de la escuela], nunca Uega tarde (Subject)
'Andrea, [who lives near the school], never arrives late1
(36) Mi hermana. [a quien conociste ayer], es doctora (Direct Object)
M v sister, [whom you met yesterday], is a doctor1
(37) Andrea, [a quien le gusta bailar], vive en Texas (Indirect Object)
'Andrea, [who likes to dance], lives in Texas'
(38) Antonio, [con quien vine ayer], es primo de Laura (Object o f Preposition)
'Antonio, [whom I came over with yesterday], is Laura's cousin'
(39) Sus amigos, [por quienes vino a la fiesta], estaban muy contentos de verla
(Circumstantial Complement (cause))
'Her friends, [because o f whom she came to the party], were very happy to see her*
(b) Que
The relative pronoun "que" can have the syntactic functions of nouns, adjectives,
or adverbs. Alarcos Llorach notes that the most frequent situation is that this relative
pronoun has an antecedent that is a noun (104). Functioning as a noun, "que" can have
the syntactic functions of Subject, Direct Object, Indirect Object, Prepositional Object,
Circumstantial Complement, and Noun Complement, illustrated below:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
19
Restrictive RCs
(40) El hombre [que trabaja aqui] es mi primo (Subject)
'The man [who works here] is my cousin1
(41) No Ie gusta la comida [que sirven aqui] (Direct Object/Nonhuman antecedent)
'She does not like the food [that they serve here]'
(42) La amiga [que traje ] es profesora (Direct Object/Human antecedent)
'The friend [that I brought] is a professor1
(43) El trabaio [al que dedica tanto tiempo] es muy bueno (Indirect Object/Nonhuman
antecedent)
'The work [to which he/she dedicates so much effort] is very good'
(44) Los estudiantes [a los que dieron buenas calificaciones] estan muy felices (Indirect
Object/Human antecedent)
'The students [to whom they gave good grades] are very happy'
(45) Los amigos [en que confiaba] no lo ayudaron en esa ocasion (Object of
Preposition)
'The friends [on which he used to rely] did not help him in that occasion'
(46) El edificio [en que ellos trabajan] es este (Circumstantial Complement/ Place)
'The building [in which they work] is this one'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 0
(47) La sociedad burguesa fde qiie1 0 form am osparte\ [...] tiende a dignificar el trabajo
(Alarcos Llorach 104).
'The bourgeois society [of which we are members]...tends to dignify work'
(48) Tiene agallas [...] para lojoven [que es] (Attribute) (Alarcos Llorach 105).
He/she is brave [...] despite the fact that he/she is so young1
In (47) the preposition and the relativizer indicate "membership" and in (48), the
antecedent is an adjective, and the relative "que" functions as an attribute. Examples like
(47) will be considered the Object o f Preposition in the present work. Only cases when
the relativizer indicates place, time, mode, cause, or means expressed by the verb will be
labeled as Circumstantial Complement, as in (46).
Nonrestrictive RCs
(49) Los candidates. [que Uegaron temprano], estaban muy nerviosos (Subject)
'The candidates, [who arrived early], were very nervous'
(50) Los libros de bioloaa. [ que Elena ha comprado], no son muy buenos (Direct
Object/Nonhuman antecedent)
'The biology books, [which Elena has bought], are not very good'
l0 In Alarcos Llorach, 'de que' was underlined (104).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 1
(51) Los amigos de Juan, [que vimos ayer], son profesores de ciencias (Direct
Object/Human antecedent)
'Juan's friends, [whom we saw yesterday], are science professors'
(52) Ese libro. [al que dedico tanto tiempo], nunca fiie publicado (Indirect
Object/Nonhuman antecedent)
'That book, [which he/she put so much time into], was never published'
(53) Esther, [a la que le escribo cartas siempre], es mi mejor amiga. (Indirect
Object/Human antecedent)
'Esther, [to whom I write letters often], is my best friend'
(54) Esas personas , [con las que comparte la oficina], son sus amigas. (Object of
Preposition)
'Those persons, [whom he/she shares the office with], are her/his friends'
(55) La mesa blanca. [sobre la que puse el libro], no es mia. (Circumstantial
Complement/Place)
'The white table, [which I put the book on], is not mine'
(c) El cual
Nonrestrictive RCs
(56) Esos articulos. [los cuales recomende a los estudiantes], son muy buenos. (Direct
Object)
'Those articles, [which I recommended to the students], are very good'
(57) Cecilia, [a la que Ie mando cartas siempre], vive en Europa. (Indirect Object)
'Cecilia, [whom I always send letters to], lives in Europe.'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 2
(58) Andrea v Elena, [con las cuales vamos al gimnasio], no quieren ir hoy. (Object of
Preposition)
'Andrea and Elena, [whom we go to the gym with], don’t want to go today.'
(59) Las habitaciones del segundo piso. [cerca a las cuales caminaban],...'
(Circumstantial Complement/Place)
'The rooms on the second floor, [near which they were walking ],...'
(d) Cuyo
Restrictive RCs
The relativizer "cuyo" is illustrated in (60) and (61):
(60) La coleea [cuyo coche fiie robado] esta hablando con la policia
'The colleague [whose car was stolen] is talking to the police'
(61) Rebeca. [cuyo coche fue robado], esta hablando con la policia
'Rebeca. [whose car was stolen], is talking to the police'
In (60) "cuyo" appears in a Restrictive RC, and in (61), the relativizer is used in a
Nonrestrictive RC.
The possessor is the person or thing that the antecedent refers to. Unlike other
relativizers we have seen, "cuyo" does not agree with its antecedent, but with the noun it
precedes within its RC (the possessed entity). This agreement is in gender and number.
It has been noted (e.g., Real Academia Espafiola; Alarcos Llorach) that "cuyo," which is a
form is used in literary and written texts, shows some variation in spoken Spanish, which
normative grammars consider a misuse. This is the case of the use of "que su" (possessive
pronoun), instead o f "cuyo." Use of the relativizer "que + su" (possessive pronoun) were
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23
more frequent than "cuyo" in the data of the present study. However, only six cases of
these structures were found in the data. The variation between "cuyo" and "que su" is
shown in (62) and (63):
(62) Aquel hombre [que su hijo esta en Africa] (Real Academia Espafiola 530)
'That man [whose son is in Africa]'
(63) Aquel hombre [cuyo hijo esta en Africa]
'That man [whose son is in Africa]'
The relative adverbs "donde," (where) "cuando," (when) and "como" (how) have
the function of circumstantial complements within their RC. They refer to notions o f
place, time, and manner or mode, respectively, expressed by their antecedent. (Alarcos
Llorach 101) These relativizers, labeled Circumstantial Complement in this dissertation,
can appear in Nonrestrictives as well as Restrictives), as shown in (64), in which "donde"
is used with a Restrictive RC:
(64) La casa [en donde nacio] es ahora una tienda
'The house [where he/she was born] is now a store'
The relativizer "cuanto" can function as an adjective, agreeing in gender and
number with the noun it modifies, as in (65). In this example, "cuantas" is qualifying the
noun "cosas," and it agrees with this noun in gender (feminine) and number (plural):
(65) Que nadie entre en nuestra escuela [que no se atreva a despreciar en si mismo
tcmtas cosas cuantas desprecia en su vecino] (Alarcos Llorach 103)
'Nobody dares to come to our school [that does not dare to despise in him/herself
as many things as the ones he/she despises in his/her neighbor]'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24
The form "cuanto" can have adverbial functions, as in (66):
(66) Influyo Scuanto1 1 pudo\ en aquel espiritu (Alarcos Llorach 103)
He/she influenced [as much as he/she could] in that spirit
"Cuanto" can also have the functions characteristic o f nouns, when the antecedent
is absent (Alarcos Llorach 103). See (67), where "cuanto" functions as a Direct Object:
(67) No respeta nada, destroza \cuanto toca] (Alarcos Llorach 103)
He/she/it does not respect anything; he/she/it destroys all [that he/she/it touches]'
In Spanish, as in English, it is possible to have the so-called existential complement
RCs. This is shown in (68):
(68) Aqui en Mexico, de las gentes mas inquietas [que habia entonces,...] (Lope Blanch,
El Habla de la Ciudad 70).
‘ Here in Mexico, out of the most active people [that existed back then,...]’
Fifty six cases o f existential RCs were found in the data o f this investigation (3345
RCs, including Human, Nonhuman, and Animal antecedents). The majority of these cases
were Restrictive RCs (53 cases).
Free relatives
In Spanish, it is also possible to use so-called free relatives. A free
relative is a RC that lacks an expressed antecedent (Campos 107). Example (69) shows
the use of a relative pronoun, "quien," as a free relative. In this example, "quien"
functions as a subject.:
“The relativizer cuanto was underlined in Alarcos Llorach (103).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25
(69) < J > [Quien no estudia] es Ana
'( { ) [Who does not study] is Ana1
The other relativizers that are used in free relatives are "cuando," "como,"
"donde," and "cuanto," illustrated in (70), (71), (72), and (73), respectively:
(70) Visitanos 4 > [cuando regreses]
'Visit us $ [whenever you return]1
(71) Baila < ] > [como quieras]
Dance $ [as you want]'
(72) Deja el libro § [donde quieras]
Teave the book C f > [wherever you want]'
(73) (Todo) < } ) [cuanto ella te dijo] es verdad
'(All) ( { ) [that she told you] is true'
The description and exemplification of the different syntactic functions for RCs
presented in this section is meant to present a clear picture o f the types of structures that
will be dealt with later in this dissertation. This facilitates the recognition of the different
syntactic possibilities, in terms of RC production in Spanish, and forms the basis of my
study of the discourse factors involved in relativization.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 6
1.2.3. The Accessibility Hierarchy for RC formation: the case of Spanish.
After studying fifty languages from a typological perspective, Keenan and Comrie
proposed that there are differences among the languages in terms o f which syntactic
positions can be relativized in the nominal phrase. As Keenan and Comrie claim: "The AH
directly reflects the psychological ease o f comprehension. That is, the lower a position is
on the AH, the harder it is to understand RCs form on that position." (88) The
Accessibility Hierarchy (AH) "expresses the relative accessibility to relativization on NP
positions in simplex main clauses." (Keenan and Comrie 66). In (74) the sign ">" indicates
"is more accessible than"; SU means "subject"; DO is "direct object"; 10 stands for
"indirect object"; OBL is a "major oblique case NP," GEN is "genitive" (or possessor),
and OCOMP means "object o f comparison" (Keenan and Comrie 66):
(74) Accessibility Hierarchy (AH)
SU > DO > 10 > OBL > GEN > OCOMP
Keenan and Comrie consider OBL "NPs that express arguments of the main
predicate, as the chest in John p u t the money in the chest rather than ones having a more
adverbial function like Chicago in John lives in Chicago or that day in John left on that
day...." (Keenan and Comrie 66)
In the present study, "in Chicago" (locative) and "that day" (temporal) in Keenan
and Comrie's examples given above would be considered Circumstantial Complement; the
expression "in the chest" would be considered circumstantial complement (place). Unlike
these authors' hierarchy, the present work includes locatives and temporals as part of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
27
Circumstantial Complement. Circumstantial Complements within RCs occur in a
considerably high number in the data (568 cases, out of 3345 cases).
I do not use the label "oblique" in this dissertation. As I indicated earlier, I
differentiate between Object of Preposition (when the relativizer is preceded by a
preposition, but excluding Direct and Indirect Object) and Circumstantial Complement
(when the complement indicates place, time, mode, cause, means, or instrument of the
action expressed by the verb (Real Academia Espafiola 375).
The AH specifies "the set of possible grammatical distinctions to which RC
formation (from simplex main clauses) may be sensitive...." (Keenan and Comrie 67)
Keenan and Comrie propose three constraints in the AH:
1. A language must be able to relativize subjects.
2. Any RC-forming strategy must apply to a continuous segment of the AH.
3. Strategies that apply at one point of the AH may in principle cease to apply
at any lower point.
The strategies that Keenan and Comrie refer to are based on the location of the
head nominal phrase (nucleus) and the Restrictive RC. In this respect, they distinguish
among three strategies. The first is postnominal, in which the nominal phrase antecedent
is placed before the RC (the case of Spanish), as in (75):
(75) El hombre [que Maria vio]
' The man [that Mary saw]1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
They also distinguish the prenominal strategy, in which the RC is produced before
the nominal phrase that it modifies, and an internal strategy, in which the RC contains the
nominal phrase.
The authors also distinguish between two different strategies of RC formation
based on the way in which the relativized position is marked within the RC. In the first
case, the authors consider [-case] the strategy in which there is no nominal element or
particle in the RC that marks the NP position that is relativized (Keenan and Comrie 65).
In (76), the [-case] strategy is illustrated for English. The strategy in (76), is [-case]
because "who" can also be used in cases in which the role of the relativized NP is
different, as in (77) and (78) (Keenan and Comrie 65-66):
(76) the girl [who John likes]
(77) the girl [who likes John]
In the second strategy, there is a nominal particle that marks the relativized
position within the RC. Such an element can be, for Comrie and Keenan, "either a
nominal element— i.e. something which has, to a significant extent, the morphological
features and syntactic distribution o f things that are clearly N P ' s in the language— or
something like a pre- or post- position, which forms a constituent with N P ' s in simple
sentences." (Comrie and Keenan 656)
The Spanish language has the following characteristics, in terms o f strategies of
relativization (Keenan and Comrie 79):
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
(78)
RC forming strategy Subj DO 10 Obi Gen OComp
1. postnom, -case + + +
2. postnom, +case -
Comrie and Keenan modified (78) in the position of IO (indirect object),
considering the use of "a quien" (preposition "a" plus relative pronoun "quien") as a
[+case] strategy (Comrie and Keenan 656). With this modification Castilian Spanish has
the following configuration, in terms of the strategies of relativization:
(79)
RC forming strategy Subj DO 10 Obi Gen Ocomp
1. postnom, -case + +
2. postnom, +case - - + + +
In example (80), from a study of relativization among six year old Mexican
children, the author of the study author found a case of an RC relativized in the position of
Subject, in which a [+case] strategy was used:
(80) Hugo [que el es el que lo trajo] (Bamga Villanueva 113)
Hugo, who is the one who brought it1
Although the example in (80) contradicts Keenan and Comrie's and Comrie and
Keenan's prediction that subjects are relativized with the strategy [-case] in Spanish, it has
to be noted that in Barriga Villanueva's data, this is the only case of the strategy [+case]
when the relativized position is Subject. Moreover, it has to be considered that (80) does
not belong to adult speech.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
The Accessibility Hierarchy (subject>direct object>non-direct object>possesor)
reflects, as Comrie (156) claimed, "ease of accessibility to relative clause formation." In
the present study, the aim is to explain patterns o f relativization, taking into account
factors involved in the discourse, which determine different frequencies for the various
syntactic roles involved in the production of a RC.
The Accessibility Hierarchy proposed by Keenan and Comrie and Comrie and
Keenan, unlike my study, is a typological study based on cognitive difficulties involved in
relativization. In Chapter 3 ,1 explain frequencies of different types of RCs, taking into
account semantic/pragmatic factors present in the interaction. The frequencies of
relativization of the different syntactic roles discussed in this dissertation are, according to
my studies, motivated by functional factors.
1.3. Linguistic variables
The syntactic role of an RC and of its antecedent result in different types of
combinations (henceforth referred to as TCs). Forty-seven different TCs were found in
the data, which is comprised of 3345 Restrictive and Nonrestrictive RCs. However, I will
only focus on the TCs that occurred most frequently. These TCs are exemplified in
Chapter 2, and Appendix 1 includes examples of each o f the other 36 TCs found in the
data. This study will analyze the correlation between the 11 TCs with semantic/pragmatic
factors such as Humanness, Grounding, Definiteness, and Functional Role of the RC with
respect to its antecedent. Each of these factors are explained in detail in Chapter 2.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
31
In what follows here, I will explain each o f the semantic/pragmatic factors whose
correlation with the 11 TCs will be analyzed in Chapter 3. I have coded the variable
Information Status following the tripartite taxonomy o f Assumed Familiarity proposed by
Prince. Prince distinguished among Evoked, Inferable and Brand-new entities in the
discourse, based on what the speaker assumes the hearer knows about the referents that
are introduced in the interaction. In Chapter 2 1 will explain in detail these three
categories and their subcategories.
The variable Information Status is important to the understanding o f the different
Functional Roles o f RCs with respect to their antecedents. The factor Functional Role of
a RC interacts with the factor Information Status o f an antecedent of an RC, as I will
show in Chapter 3.
The Information Status or Informational Value of the information conveyed by the
RCs is also worth taking into account in the analysis of the various functions of the RCs in
the present study.
I follow Fox and Thompson's definition o f Grounding: "To ground a noun phrase
is to locate its referent in conversational space by relating it to a referent whose relevance
is clear, that is, to a Given referent in the immediate context." (300)
The three types o f Grounding described by Fox and Thompson are Anchoring
(300), Main-Clause Grounding (300-01), and Proposition-Linking Grounding (301).
These three types of Grounding represent three different strategies that speakers use to
make referents relevant in the interaction. They are important to consider here because of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
32
the possible correlations between types o f Grounding and the distribution of TCs in
Spanish.
Fox and Thompson identified two different types o f Functional Roles in their study
of RCs in conversational English. One of the functions was Characterization o f an
antecedent "not previously known to the hearer." (301) The other function consisted of
helping to identify the referent of the antecedent of an RC (302). In the present study,
four different functional roles were analyzed: Characterization of an Evoked, Brand-New
or Inferable antecedent, Giving New information about an Evoked, Brand-New or
Inferable antecedent, Identifiyng an Evoked antecedent, and Reaffirming the existence of
an Evoked or Inferable antecedent.
I will show that Humanness of a noun phrase antecedent and Grounding interact
with the different types o f functional roles that the RCs can have in Spanish.
The factors Humanness and Definiteness will be relevant in this investigation for
the explanation of the patterns of distribution of the syntactic roles that are involved in
relativization. The factor Humanness of the antecedent o f an RC helps to explain patterns
of relativization but it is also useful for the explanation o f the different types of Grounding
or relevance that a referent can receive in the discourse.
The factor Definiteness will also be taken into account in this study in the analysis
of Functional roles that the RCs can have in the discourse.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33
1.4. Social variables
The present study analyzes the use of Restrictive and Nonrestrictive RCs and the
variable use of a preposition with the relativizers "que/that" and "donde/where" in contexts
in which this variation is possible. Level of education and socioeconomic level are factors
that will be crucial in the explanation of the distribution o f these variables. In (81) and
(82) I illustrate the variable use o f prepositions with the relativizers "que" and "donde":
(81) ...hasta que llego un momento [en que dije:...] (Lope Blanch, El Habla de la
Ciudad 17)
‘..until a moment arrived [in which I said:...]’
(82) En eso, al voltear al salon [donde yo tomaba clases a esa hora], vi que... (Lope
Blanch, El Habla de la Ciudad 17-18).
'Then, after turning to the classroom [in which I was taking classes at that time], I
saw...'
In (81), it is possible for the preposition ‘en/in’ to be omitted, and in (82), it is
possible for a preposition to be produced before the relativizer ‘donde/where’.
The theoretical construction of social class has been researched. The concept
"social class" has been used to stratify speakers in communities in previous sociolinguistic
studies. A typical example is Labov's identification of the variable (th) in New York City
as a sociolinguistic marker because the frequencies of its variants correlate with social and
stylistic factors (Sociolinguistic Patterns 112-3). Labov placed the speakers in his study in
different social classes by calculating a socioeconomic index, considering three indicators:
(1) the occupation o f the person who was responsible o f supporting the household, (2)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
34
education o f the speaker and (3) the household's income ('Sociolinguistic Patterns 112-3).
The social levels analyzed by Labov were: lower class, working class, lower middle class,
upper middle class ('Sociolinguistic Patterns 112-3).
The term "socioeconomic factor" is frequently used in sociolinguistic research, and
it is defined using objective criteria such as income, profession, level of education, location
and type o f residence. The use of these criteria in sociolinguistics allows the investigators
to separate speakers into several levels that are defined as "discrete groups," although it is
understood that "social class" and the criteria used to define it are indeed "relative social
continua" (Silva-Corvalan, Sociolinguistica 78).
As Silva-Corvalan notes, the parameter "level o f education" is an effective
predictor o f the linguistic behavior o f speakers (Sociolinguistica 79-80). The prescriptive
grammar that is taught to children and teenagers creates an awareness about the non-
prestigious linguistic features. Moreover, the continuous exposure to standard and written
language facilitates self correction and the suppression of nonstandard forms.
There are several works that have indeed demonstrated the existence of social
differences in the use of RCs. One such study is Macaulay's, which found social
differences between two social groups in the use o f Restrictive and Nonrestrictive RCs in
his interviews in Ayr, Scotland.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
35
Macaulay found differences between the two social groups he studied (middle
class and lower class). Among other findings, the author noticed that Nonrestrictive RCs
were more frequently used among the speakers o f the middle-class (Macaulay 64).
Kroch and Hindle, studying Philadelphia English, stratified their subjects in three
social groups, upper-class adults, lower middle-class college students, and working-class
adults, and examined the occurrence o f RCs among these groups (Kroch and Hindle 69).
The upper class (UC) subjects showed more frequency in their use o f RCs than the lower
middle class students (LMC) and working class adults (WC) (Kroch and Hindle 69, cited
by Finegan and Biber 331-32). The speakers from the upper-class produced 14 RCs per
100 sentences, whereas the lower middle-class students and the working- class adults
showed a lower frequency o f use of RCs (5 RCs per 100 sentences and 3 RCs per 100
sentences respectively).
Lindenfeld noticed, in a French-speaking community, that professional speakers
showed the use of more relativization than speakers from the working class group (cited at
Biber and Finegan 331).
Van den Broeck's study of Maaseik, a Flemish-speaking community, includes
relativization. Middle class speakers showed, in two situations of use and in both formal
and informal contexts, longer T-units1 2 than the subjects that belonged to the lower class
(Finegan and Biber 333). In Chapter 4 1 will establish correlations of the use of
1 2 As Van den Broeck explains, the concept o f T-unit has been developed by
Hunt (Van den Broeck 155). It can be defined as "one main clause plus the
subordinate clauses attached to or embedded within it".
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
Restrictive and Nonrestrictive RCs and the variables level of education and socioeconomic
level.
The levels of education of the speakers in the present investigation are: complete
primary education; incomplete primary education; complete secondary education;
incomplete secondary education; technological school; complete college education; and
incomplete college education.
In this study, with the information provided by the transcriptions o f the data that
were used, the speakers were classified in two socioeconomic levels, taking into account
the speakers’ profession (which implies a level o f education and often indicates income
and standard of living as well). I will also consider information provided by the speakers
during the interviews that may further indicate their standard of living, such as trips, social
and professional activities, etc.
The two socioeconomic levels were middle and lower. The division between
middle and lower levels allows me to distinguish between a philosophy professor and a
janitorial employee.
As I have stated above, the division between middle and lower socioeconomic
levels took into account information in the transcribed corpora before each conversation
and what the speakers themselves revealed about their socioeconomic situation during the
course o f the conversations. Because already transcribed data was used, this approach
was necessary.
Finally, a factor that has to be considered in the present study is the geographical
origin o f the speakers. The speakers are included from three different transcribed corpora:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
El Habla de la Ciudad de Mexico (corpus of middle class speakers from Mexico City); El
Habla Popular de la Ciudad de Mexico (corpus o f lower class speakers from Mexico
City); and El Espanol Hablado en el Suroeste de los Estados Unidos (corpus o f the
Spanish spoken in southwestern United States). By including the speakers from the
southwestern United States in the present investigation I am able to investigate the
influence of these speakers' lower educational level in Spanish on their variable use of
prepositions with a relativizer.
The speakers in El Habla de la Ciudad de Mexico and El Habla Popular de la
Ciudad de Mexico were bom in Mexico City, and have lived in that city all their lives.
Actually, a few speakers were not bom in Mexico City but moved there at a very early
age. The speakers in El Espanol Hablado en el Suroeste de los Estados Unidos were bom
in one o f four different places: San Marcos, Texas; Mora, New Mexico; Tucson, Arizona;
and San Jose, California. The parents and, in most cases, the grandparents of these
speakers were also bom in these regions.
Because they were exposed to both English and Spanish at a very early age, these
speakers constitute, in terms of their use of Spanish, a different group than both El Habla
de la Ciudad de Mexico and El Habla Popular de la Ciudad de Mexico. I will investigate
the possibility that the speakers in El Espafiol Hablado en el Suroeste de los Estados
Unidos differ from the speakers in El Habla de la Ciudad de Mexico and in El Habla
Popular de la Ciudad de Mexico, in their use of a preposition with a relativizer. Because
they use Spanish primarily at home or in very limited situations, it is possible that these
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
bilingual speakers will show a less frequent use o f the standard variant, that is, the
presence of a preposition with a relativizer when its absence is also possible.
1.5. Situational variables
1.5.1. Register
In the present study, the terms "Register" and "Register variation" are used,
following Finegan and Bibefs definition, "as cover terms for the full range o f language
varieties associated with differences in communicative situation (including mode and
purpose)." (316)
The variable Presence/Absence o f a Preposition is analyzed across various types of
registers here in order to explain the influence different communicative situations have on
the distribution o f its variants. I will analyze the following types o f registers:
1. One-to-one interview (directed dialogue).
2. One-to-one interview (free conversation).
3. One-to-one interview (secret recording).
4. One-to-one interview (secret recording, with a third person that participates a few
times)
5. One-to-one interview (directed dialogue with the presence of a third person who speaks
occasionally).
6. One-to-One interview (free conversation, with a third person who participates)
7. Group interview (directed dialogue, interviewer participates).
8. Group interview (free conversation, interviewer participates).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
39
9. Group interview (directed dialogue, interviewer participates, and a third person
participates).
10. Group interview (free conversation, interviewer does not participate).
11. Group interview (secret recording, interviewer participates).
12. Group interview (free conversation. Interviewer participates and third person also
participates).
13. Group interview (free conversation. Interviewer participates and a third person
participated twice ).
14. Monologue.
15. Professional conferences.
I will show that professional conferences are the register that favors the presence
of a preposition. The fact that these conferences are about academic topics may influence
the more frequent use o f the standard variant (presence o f a preposition) with respect to
the one-to-one or group conversations.
1.6. Summary
In the previous sections I have presented the semantic/pragmatic, social, and
situational variables that will be taken into account in the analysis o f the grammar o f RCs
in Spanish.
Chapter 2 explains the methodology used in this dissertation. Chapter 3 presents
and discusses the results o f the analysis of the linguistic variables, Chapter 4 examines the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
results of the analysis of social and situational variables, and Chapter 5 presents the
conclusions o f this investigation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2
41
METHODOLOGY
2.1. The speakers and the data
This study analyzes data from Mexico City and from the southwestern United
States. This data includes speech samples from a total of 98 speakers, men and women,
distributed among the following age groups:
a. 17-35 years old
b. 36-55 years old
c. 56+ years old
A total of 79 o f the speakers are from Mexico City and have lived there for most
o f their lives. The remaining 19 speakers are from the southwestern United States,
specifically from San Marcos, Texas; Mora, New Mexico; Tucson, Arizona; and San Jose,
California.
The two socioeconomic levels in which the speakers were classified in this
investigation are the following:
a. Middle class: individuals who have professional careers: professors, accountants,
doctors, etc.
b. Lower class: individuals who work as domestic employees, security guards, cleaning
employees, etc.
In the case of housewives, the source of income of the total household was
considered in order to classify the speakers in a socioeconomic level.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
The data were obtained from the following sources:
a. El Habla de la Ciudad de Mexico (edited by Lope Blanch). 24 transcriptions o f 30-60
minute recordings from this publication were used, and they include the following types of
interviews: 11 one-to-one interviews between a researcher and a subject; ten dialogues
between two speakers (some of these interviews are directed dialogues);one monologue,
and four conferences.
b. El Habla Popular de la Ciudad de Mexico (edited by Lope Blanch) includes 34
transcriptions of 30 minute recordings of the following types: 15 recordings corresponding
to one-to-one interviews between a researcher and the speaker; 11 recordings o f dialogues
between two speakers (in all of them the interviewer participates occasionally); 8
transcriptions corresponding to secret recordings1 3 (6 of these interviews are one-to-one
and 2 are group interviews (two speakers).
c. El Espanol Hablado en el Suroeste de los Estados Unidos (edited by Lope Blanch).
Contains transcriptions of 30 minute one-to-one interviews between an interviewer and a
speaker. There are 19 interviews in this corpus.
The main topics discussed in the interviews were the speakers' occupations, past
experiences, and daily activities. The conferences were about specific academic topics
pertaining to the speakers' professions.
1 3 It was not specified in Lope Blanch (El Habla Popular) whether these
subjects were told or not that there had been a secret recording after it took place.
However, these recordings were transcribed and the transcriptions were made
available to the public.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
The statistical program SPSS 7.5 for Windows was used for the quantitative
analysis in the present study, specifically crosstabulations. In crosstabs, the Pearson chi-
square is a statistic that tests whether or not the row and column variables are independent
or not. The Pearson chi-square "is calculated by summing over all cells the squared
residuals divided by the expected frequencies." (Norusis 207)
In an example in which the Pearson chi-square value of a correlation is 22.56, and
the observed significance level is .00001, if the two variables are independent, "the
probability that a random sample would result in a chi-square value of at least that
magnitude is 0.00001." (Norusis 208)
If the observed significance level o f the test is less than 0.05 or 0.01, the
hypothesis that two variables are independent will be rejected. (Norusis 208).
2.2. Coding for analysis of Linguistic Factors
This study incorporates several semantic/pragmatic and social factors. The main
variable is the set o f possible combinations of grammatical roles involved in the production
o f relative clauses (RCs in this study). The grammatical roles of the RC and of its
antecedent result in different possible Types of Combinations (called TCs in this study). A
total of 47 TCs were found after the coding of the data, which consists of 3345 Restrictive
and Nonrestrictive RCs (examples for each of these types of combinations are shown in
Appendix 1).
I will focus only on the 11 TCs that were most frequent in the data. I will use the
phrase 'A-B' to represent a phrase in which the antecedent o f the RC has the grammatical
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
role 'A' and the relativized position has the grammatical role 'B.1 Following Fox and
Thomson (1990: 298), these 11 TCs are as follows:
1. Circumstantial Complement-Circumstantial Complement (cc-cc):
Hay cinco habitaciones en la casa [en que mi familia vive]
'There are five rooms in the house [where my family lives ]'
2. Circumstantial Complement-Subject (cc-s):
Estuvimos en la casa [que tiene cinco habitaciones]
W e were in the house [that has five rooms]1
3. Object of Preposition-Subject (op-s):
Fuimos con la amiga [que vino de Toronto]
W e came over with the friend [that came from Toronto]'
4. Object of Preposition-Direct Object (opdo):
Fuimos con una amiga [que invitamos a ultima hora]
W e went with a friend [that we invited at the last minute]1
5. Predicate Nominal-Subject (pn-s):
Ella es una estudiante [que tiene buenas calificaciones]
'She is a student [that has good grades]'
6. Predicate Nominal-Direct Object (pn-do):
Ella es la estudiante [que aceptamos este aiio]
'She is the student [that we accepted this year]'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7. Direct Object-Subject (do-s):
Traje el libro [que tiene articulos de Literatura]
'I brought the book [that has articles on Literature]'
8. Direct Object-Direct Object (do-do):
Traje el libro [que compre ayer]
'I brought the book [that I bought yesterday]'
9. Subject-Subject ( s-s ):
Las personas [que asistieron a la reunion] presentaron sus trabajos
'The persons [that attended the meeting] presented their works'
10. Subject-Direct Object (s-do):
Las computadoras [que compramos] funcionan perfectamente
'The computers [that we bought] work perfectly'
11. Existential-Subject (ex-s):
Hay un hombre en la oficina [que habla cinco idiomas]
'There is a man in the office [that speaks five languages]'
I will analyze the correlation o f the 11 TCs shown above with the factors
Humanness, Grounding, Information Status, Definiteness, and Functional Role o f the RC.
These variables will be analyzed in this study in terms of their correlation with the
frequencies of distribution of the 11 TCs that were shown above: 'cc-s,1 'cc-cc,' 'op-s,' 'op-
do,' 'pn-s,' 'pn-do,' 'do-s,1 'do-do,' 's-s,' 's-do,' and 'ex-s.'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
46
2.2.1. Restrictive vs Nonrestrictive relative clause:
The present study makes a distinction between Restrictive and Nonrestrictive RCs.
Fox and Thompson decided not to make this distinction because, as they claim: "on
intonational grounds, we found no clear cases o f nonrestrictive relative clauses in our
conversational data, and several cases which were indeterminate." (297-8)
As I indicated in Chapter 1, the criteria used to distinguish between Restrictive and
Nonrestrictive RCs are the following:
a. A Restrictive RC specifies and restricts the set of referents in an antecedent (Real
Academia Espanola; Bello; Campos), and is "crucial" for the identification of an
antecedent (Campos 98).
b. A Nonrestrictive RC merely provides additional information about a referent that has
already been specified or restricted. (Real Academia Espanola; Bello; Campos).
Although the transcriptions of the data used in this investigation use commas to
mark the separation between a nominal phrase antecedent and its RC in Nonrestrictive
RCs, the criteria recapitulated above were taken into account in order to determine
whether a RC is Restrictive or Nonrestrictive.
The main characteristic of a Nonrestrictive RC o f not being crucial to specify a
referent is equivalent to what Klein-Andreu observed in noncontrastive adjectives.
According to Klein-Andreu, a noncontrastive (preposed) adjective may be interpreted as a
comment (162), as can be observed in (83):
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
47
(83) Para muchos, esta inversion es un derroche.
"Costara al pais 400 millones de pesetas para obtener un triste megavatio de
potencia,... (Klein-Andreu 162).
Tor many, this investment would be a waste. "It will cost the country 400 million
pesetas to obtain one sorry megawatt...' (Klein-Andreu 162)
In (83), the adjective "triste/sorry" is noncontrastive in Spanish because it does not
establish a distinction. The adjective "triste/sorry" is merely qualifying the noun
"megavatio/megawwatt." This noncontrastive feature of preposed adjectives like
"triste/sorry" in (83) is comparable to the most distinctive feature o f Nonrestrictive RCs,
which is the fact that they typicallly provide a comment about an antecedent rather than
specifying or restricting it. If the adjective "triste/sorry" is postposed, it will no be the
case that it conveys a negative comment about a noun in Spanish. This can be observed in
(84):
(84) una pelicula triste
'a sad film1
In (84), the postposed adjective "triste/sad" does not have the interpretation of a
negative comment about a noun, as was observed in (83). In (84), the interpretation of
the postposed adjective is restrictive, since it restricts or specifies the reference of the
noun "pelicula/film."
Discussing adjective placement in Spanish, Klein-Andreu claims that, "the greater
the overall intent to comment, rather than simply to describe fact, the higher the
proportion we should expect of preposed characterizations." (Klein-Andreu 171). In the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
48
present study, the same can be said about the use of Nonrestrictive RCs: the greater the
overall intention to provide a commentary about an already specified antecedent, the
higher the frequency of use o f Nonrestrictive RCs by the speakers.
The typical role o f Nonrestrictive RCs, that of providing a commentary about an
already specified antecedent instead of having a restricting role, can be characterized as
being evaluative. Evaluation is one of the various components of a narrative in Labov's
(1972b) definition of a narrative passage. Labov considers a narrative " a sequence of two
clauses which are temporally ordered." (Labov, Language 360). A narrative may have
several parts: abstract, orientation, complicating action, evaluation, resolution, and coda
(Labov, Language 363).
The evaluation indicates "the point of the narrative," the reason for which it was
told. An evaluation explains the point of the narrative (Labov, Language 366).
Nonrestrictive RCs tend to have an evaluative function, that is, they tend to convey
the speakers' opinion or view o f an antecedent.
In Chapter 4 ,1 will show the correlation of the use of Restrictive and
Nonrestrictive RCs with social variables such as level o f education and socioeconomic
level. That analysis will explain that the higher frequency o f use of Nonrestrictive RCs
among speakers from a social group may indicate the existence o f different styles of
communication, at least with respect to relativization.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
49
2.2.2. Humanness
This variable contains two possibilities in terms o f the nominal phrase antecedent
being either Human or Nonhuman. A third category for Animals only was also considered
in the coding o f this variable, but it will not be considered in the quantitative analysis in
Chapter 3 due to the limited number of RCs with Animal antecedents (33 cases) in the
data analyzed in this investigation.
2.2.3. Definiteness
This variable includes the following categories for the noun phrase antecedents o f
the RCs: Definite, Indefinite and Noncountable (for mass nouns).
For the definition of a Definite noun phrase, I follow Fox and Thompson, who
consider a definite noun phrase antecedent “any noun phrase occurring with a definite
article or demonstrative, as well as nouns with possessives.” (301)
According to this definition, the noun phrase in (85) was coded as Definite:
(85) El hiio mayor, [que va a tener treinta y seis anos], el es el que corre todo el
negocio (Lope Blanch, El Espanol 243)
'The oldest son, [who is going to be thirty six years old], he is the one that runs the
business'
In cases in which the nominal phrase is not explicit, the definiteness or
indefiniteness of the antecedent is also taken into account. According to this, the
antecedent in (86) was coded as Definite because the article "la" is Definite:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50
(86) ...a la [que quise mas], nunca fiie mi novia. (Lope Blanch, El Habla Popular 376)
'...the one [that I loved the most], she was never my girlfriend.'
If the following Indefinite elements precede a noun phrase antecedent, such
antecedent was considered Indefinite: alguno/a ('some' masculine/feminine ), algunos/as
(the plurals of alguno/a) , ninguno/a ('any' masculine/feminine), ningunos/as (the plurals o f
ninguno/ninguna), cualquiera ('any' or ’ either1 ); quantitative identifier todo (everything)
and its variations o f gender and number, mas (more), menos (less), mucho ('much' or 'too
much') and its variations o f gender and number, and poco (little') and its variations of
gender and number, among others (Real Academia Espanola 230-36).
A third category in this variable is Noncountables, in order to classify nominal
phrases that are 'mass nouns.' Of the 3345 Restrictive and Nonrestrictive RCs in my
study, only 46 antecedents were Noncountables.
2.2.4. Information Status of Nominal Phrases Antecedents of the Relative Clauses
This variable was coded using the tripartite taxonomy of Assumed Familiarity
proposed by Prince. She proposed a scale of Assumed Familiarity using the analogy of the
writing of recipes to illustrate how a writer verbalizes a recipe depending on his or her
"assumptions about what the hearer knows about ingredients, processes, and equipment,
about what equipment the reader has available, and about what staples the reader keeps on
the shelf." (Prince 234-35) According to Prince, a text constitutes " a set of instructions
from a speaker to a hearer on how to construct a particular DISCOURSE MODEL."
(235). A discourse-model object, that is, a discourse entity, can represent “an individual
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(existent in the real world or not), a class of individuals, an exemplar, a substance, a
concept, etc.” (235).
Prince (237) proposed the scale of Assumed Familiarity shown in Figure 1:
A ssum ed Familiarity
In ferra b le ^ Evoked
Brand new Unused (Noncontaining) Containing (Textually) Situationally
Inferrable Inferrable Evoked Evoked
Brand-new Brancf-new
(Unanchored) Anchored
Figure 1
The examples in (87) were used by Prince (233) to illustrate the different entities
her scale of Assumed Familiarity.1 4
1 4 In Prince, the word Inferable is spelled "Inferrable". In the present
investigation I use Inferable, with the exception of citations from Prince's text.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
52
(87) a. Pardon, would you (Situationally Evoked1 5 ) have
change o f a quarter?
b. Noam Chomsky (Unused) went to Penn.
c. I got on a bus (Brand-new Unanchored) yesterday and the driver
(Inferable) was drunk.
d.. A guy I work with (Brand-new Anchored) says he
(Textually Evoked) knows your sister.
e. Hey, one of these eggs (Containing Inferable) is broken!
Prince defines "New" in the following way: “[w]hen a speaker first introduces an
entity into the discourse, that is, tells the hearer to 'put it on the counter’ we may say that
it is NEW.” (235). She distinguishes between two types of new discourse entities. The
first is the "Brand-new," when “the hearer may have had to CREATE a new entity”
(Prince 235). The other discourse entity, "Unused," occurs when the hearer “may be
assumed to have a corresponding entity in his/her own model and simply has to place it in
(or copy it into) the discourse model, akin to taking some staple off the shelf, when its
presence is suddenly taken for granted in a recipe (e.g., salt).” (Prince 235)
Prince also notes the difference between Brand-new entities that are "Anchored"
and brand-new entities that are "Unanchored" (236). She states, "[a] discourse entity is
Anchored if the NP representing it is LINKED, by means of another NP, o r ' Anchor,’
lsThe insertions between parentheses in 88 (a-e) did not exist in Prince's
examples (233).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53
properly contained in it, to some other discourse entity" (236). A Brand-new entity in
(87(d)) is "a guy I work with," which contains the noun phrase "I." Prince explains, “the
discourse entity the hearer creates for this particular guy will be immediately linked to
his/her discourse entity for the speaker.” (236) On the other hand, a Brand-new
Unanchored entity in (87(c)) is 'a bus.1
Prince notes in her data that all entities that are Anchored "contain at least one
Anchor that is not itself Brand-new." (236) According to this observation, it is possible to
find noun phrases such as (88(a)), (88(b)), (88(c)), and (88(d)), but a case like (89) is not
possible (236).
(88) (a) I
I (b) John |
a guy— ! (c) the plumber | — work(s) with
(d) a woman I know
(89) ?a guv [a woman works with]
In each o f the examples in (88), we are able to find at least one entity that is not
Brand-new (i.e., I, John, the plumber, and I in (a), (b), (c), and (d), respectively). On the
other hand, in (89), both the noun phrases a guv and a woman represent brand-new
entities.
Notice that it is possible to have noun phrases that are Unused and anchors, as in
(90):
(90) a guv [President Clinton works with]
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
54
In (90), the Brand-new entity a guv is anchored by the Unused entity President
Clinton, which is assumed to be in the hearer’ s model. This Unused entity within the RC
"[President Clinton works with]” anchors the Brand-new entity "a guy." In (89), the RC
"[a woman works with]" contains the Brand-new entity "a woman." which cannot anchor
the antecedent "a guy."
Continuing with Prince’s Assumed Familiarity, she defines 'Evoked' entities in the
following way: “[n]ow, if some NP is uttered whose entity is already in the discourse-
model or "on the counter," it represents an EVOKED entity." (236). Prince distinguishes
between two types of Evoked entities: Textually Evoked and Situationally Evoked.
Entities are Textually Evoked when the hearer had evoked an entity earlier, “on textual
grounds, by following instructions from the speaker.” On the other hand, entities are
Situationally Evoked when the hearer evoked the entity by himselfTherself, ‘Tor situational
reasons” (Prince 236). In (87(d)), "he" is an example of a Textually Evoked, and "you" in
(87(a)) is a case of a Situationally Evoked referent.
Prince also considers entities called "Inferables": “[a] discourse entity is Inferrable
if the speaker assumes the hearer can infer it, via logical- or, more commonly, plausible-
reasoning, from discourse entities already Evoked or from other Inferrables.” (Prince 236)
In example (87(c)), the referent of'the driver1 is inferable from "a bus", "plus assumed
knowledge about buses, that is, Buses have drivers." (Prince 236). The so-called
"Containing Inferrables", where what is inferenced off of is properly contained within the
Inferrable NP itself, is illustrated in (87(e)): "one o f these eggs is a Containing Inferrable,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
55
as it is inferrable, by a set-member inference, from these eggs, which is contained within
the NP and which, in the usual case, is Situationally Evoked." (236)
In this investigation, I use the labels Brand-new, Unused, Evoked, and Inferable. I
do not distinguish between containing and Noncontaining Inferable entities, textually and
Situationally Evoked entities and Brand-new Unanchored and Anchored entities, in order
to maintain a sufficient number o f cases for each category in the quantitative analysis.
(91) is an example of with a noun phrase antecedent that is Brand-new:
(91) Entonces, (el) encontro una muchacha [que era huerfana tambien de papa y
mama.] (Lope Blanch, El Habla Popular 22).
'Then, he found a girl [that was also an orphan.]'
In (91), the nominal phrase 'una muchacha/a girl' is a Brand-new referent for the
hearer in the context of the conversation.
The example in (92) shows a case o f an antecedent coded as Inferable, because it is
assumed that the hearer is able to logically infer the referent of the noun phrase
antecedent, based on the preceding context:
(92) ..un dia acudi a la escuela en la manana, y supe que Ios companeros, el dia anterior,
habian decidido no ir a clases hasta despues de Semana Santa. En eso, al voltear
al salon [donde yo tomaba clases a esa hora], vi que estaba el maestro esperando a
los alumnos. (Lope Blanch, El Habla de la Ciudad 1)
'...one day I went to school in the morning, and I knew that my classmates, the day
before, they had decided not to go to classes until after Easter. Then, after turning
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
56
to the classroom [in which I was taking classes at that time], I saw the teacher
waiting for the students.'
In (92), the referent o f the nominal phrase "salon" can be inferred by the hearer
from the previous mention of "la escuela/ school," assuming that the hearer is able to infer
that schools have classrooms.
Unused referents are New in the discourse, but they could be assumed to be
known by the hearer. Consider the example in (93):
(93) Posteriormente se puso la Normal, [donde curso Olivia.] (Lope Blanch, El Habla
de la Ciudad 79)
E ater on, the Normal, [where Olivia studied], was instituted'
*La Normal' is an institution that can be assumed to be known by the hearer in
Mexico City. For this reason I coded this nominal phrase as Unused.
In (94), a nominal phrase was coded as Evoked. In this example, "esa fecha/that
date" is Evoked because it clearly refers to the previous mention o f the date of the
competition given by the speaker, that is, "mil novecientos cincuenta y ocho/nineteen fifty
eight."
(94) En el afio de mil novecientos cincuenta y...ocho, o cincuenta y siete, se hizo una
competencia de selection para ir al dual meet (mit) en Cuba...A partir de esa fecha.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
57
[en que quede ya seleccionado para ir a Cuba, al dual m eet...] (Lope Blanch, El
Habla de la Ciudad 13)
'In the year of nineteen fifty ...eight, or fifty seven, a competition was done to go
to the dual meet in Cuba...Since that date, [in which I was already selected to go to
Cuba, to the dual meet...]'
2.2.5. The variable Grounding
The factor Grounding is, as will be shown further in this investigation, crucial
among the semantic/pragmatic factors that I investigate as correlating with the variation of
syntactic roles within the RCs and in their antecedents.
I follow Fox and Thompson, who define Grounding in the following way: ‘T o
ground a noun phrase is to locate its referent in conversational space by relating it to a
referent whose relevance is clear, that is, to a Given referent in the immediate context.”
(300)
They considered three types of Grounding: Anchoring, Main-Clause grounding,
and Proposition-Linking grounding. The first type is Anchoring. Fox and Thompson
(300) used Prince's (236) definition o f Anchoring: “A discourse entity is Anchored if the
NP representing it is LINKED, by means of another NP, or 'Anchor.’ properly contained
in it, to some other discourse entity." The RC contains a Given1 6 referent that gives
1 6 ,Given', according to Fox and Thompson (300) is: “A referent presumed
to be in the hearer's focal consciousness (Chafe, "The Deployment"; Chafe,
"Cognitive").” In the present investigation, I use the term 'Evoked' (Prince).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
58
relevance to an antecedent introduced in the main clause. (95) shows a case in which an
RC anchors its antecedent:
(95) (talking about the upkeep of houses)
But uh- the original price of it, eh- you can't even (inaud.) the
original price,
just that little screen porch alone is five hundred dollars,
the air condish- the uh heater thins: [we put in] I think was a hundred
uh five six hundred dollars (Fox and Thompson 300).
Because o f the role that the speaker has as a participant in the interaction, 'we' is a
Given referent. The noun phrase antecedent 'the heater thing' is anchored by the RC. The
Given referent within the RC is related to the antecedent "the heater thing" (Fox and
Thompson 300).
(96) shows an example coded as Anchoring in my data:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
59
(96) Bueno, pues, de mis estudios, como usted sabe, fiieron estudios eclesiasticos,
fundamentalmente. Claro, despues de ia secundaria v...esos estudios de comercio
[que (yo)1 1 tuve necesidad de hacer despues de salir de la secundaria.] (Lope
Blanch, El Habla de la Ciudad 55).
Well, about my studies, as you know, they were religious studies, basically. O f
course, after high school and.. .those business studies [that I had the chance to do
after finishing high school.]1
In (96), the nominal phrase antecedent 'esos estudios de comercio/those business
studies’ has not received Proposition-Linking grounding, that is, it does not have a
semantic link to a proposition in the immediately anterior context. As can be observed,
there is not Main-Clause grounding in the main clause either. By means of linking the
nominal phrase antecedent with the implicit Evoked referent "yo/I", which refers to the
speaker, the RC the one that performs the Anchoring, by means of linking the nominal
phrase antecedent with the implicit Evoked referent "yo/I."
Although it is very common for the Anchoring elements to be personal pronouns
referring to one of the participants in the interaction, Anchoring elements and grounding
elements in the case of Main-Clause grounding may be any Evoked referent from the
earlier discourse.
1 7 I indicate implicit subjects or other categories by inserting them in the
examples between parentheses and in italics. These italicized insertions were not
originally in the transcriptions that I used.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6 0
The second type of Grounding defined by Fox and Thompson, Main-Clause
grounding, occurs when an Evoked (Given in these authors’ terminology) referent in the
main clause and the referent o f the noun phrase antecedent o f an RC are linked (Fox and
Thompson 300-01). With Main-Clause grounding, Fox and Thompson notice that the
noun phrase being grounded in the main clause tends to be an Object, as in (97):
(97) he's got-a spring [that comes, way up], (Fox and Thompson 301)
A Given referent in the main clause Che1 ), which is the subject o f the main clause,
grounds the object "a spring," and the RC '[that comes, way up]' does not perform any
Grounding (Fox and Thompson 301). Fox and Thompson noticed in their data that in
these cases the RC performed other functions with respect to its antecedent, and that the
most common one was Characterization (305).
Fox and Thompson usually found a "semantically neutral main verb expressing
possession, such as have or has got (Givon 1979)" in the main clause in cases of Main-
Clause grounding (Fox and Thompson 300).
They also noticed that with Existential Fleads, there is a tendency to have Main-
Clause grounding by means of using locative expressions, as in (98):
(98) there were two people THERE [who were constantly on stage] (Fox and
Thompson 309)
In order to be consistent with the definition o f Main-Clause grounding, and of
grounding in general, it has to be assumed that the locative expression THERE constitutes
a Given referent in the immediate context.
In (99) we can see an example o f Main-Clause grounding:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
61
(99) Entonces, tuve que ir viendo cual era mi defecto en los ultimos veinticinco metros
para (yo) ir mejorando y superando ese problema [que tenia]. (Lope Blanch, El
Habla de la Ciudad 14)
'Then, I had to observe what my defect in the last twenty five meters was, in order
to improve and get over that problem [that I had]1 .
Earlier in the conversation, the speaker had explained that at a certain point during
his swimming competitions, he would get tired. The antecedent "ese problema/that
problem" refers to that specific problem that he had in his competitions.
In (99), the noun phrase antecedent "ese problema/that problem" is grounded by
the implicit personal pronoun "yo/I," contained in the phrase "para ir mejorando y
superando ese problema/in order to improve...." This example is different than the one
observed in (96), in which the antecedent 'esos estudios de comercio/those business
studies' depended on the RC to obtain Anchoring by an Evoked referent within the RC. In
(99), on the other hand, by the time the RC was produced by the speaker, Main-Clause
grounding had already occurred. The production o f an antecedent of a RC and of its RC
occur in a linear order, and following Fox and Thompson, "there is no communicative
need to ground a referent more than once." (305)
The third type of Grounding analyzed by Fox and Thompson is Proposition-
Linking (301). One of the examples presented by Fox and Thompson is (100):
(100) The mother’s sister is a real bigot. Y’know and she hates anyone [who isn’t a
Catholic] (Fox and Thompson 301)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6 2
In (100), according to Fox and Thompson, the RC "[who isn’ t a Catholic]," along
with its antecedent 'anyone1 are grounded by their link to the proposition (the mother’s
sister is a real bigot), in Fox and Thompson’s words, "through the frame invoked by bigot"
(Fox and Thompson 301).
In (100), Grounding occurred in an RC and its antecedent made relevant by their
link to a preceding proposition.
A second example presented by Fox and Thompson of Proposition-Linking
grounding is a case in which an RC is linked to an earlier proposition. This is shown in
( 101):
(101) B: Y’know I’ve been reading about very old people lately, (0.4)
A: Yea//: h?
B: Like they had an article in the Rolling Stone with this guy who’s supposed
to be a hundred and thirty. The oldest American. He’s a black guy who
lives in Florida and they interviewed him ,...
B: and one thing they said in the article that was really intriguing was, in the
United States at this point, there are over a hundred, thousand people [who
are over a hundred years old] (Fox and Thompson 309)
In (101), the RC "[who are over a hundred years old]," by its link to the
proposition "Y’know I’ ve been reading about very old people lately," grounds its
antecedent "over a hundred thousand people."
The example in (102) illustrates a case of Proposition-Linking grounding from my
data:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
63
(102) Enc.- A ver. Cuenteme esto de sus hijos, de su familia...
Inf.-...mi esposo y yo comenzamos a...en mil novecientos cincuenta, a un negocio
de techeria, y aam...comenzamos sin nada,...Todos los ...trabajan ahi. El hiio
mayor, [que va a tener treinta y seis anos],...es el que corre todo el negocio...'
(Lope Blanch, El Espanol 243)
Enc.-Let's see. Tell me about your siblings, about your family...
Inf.-...my husband and I began in...in nineteen fifty, in a roofing business, and we
began from scratch,... All o f them... work there. The oldest son, [who is going to
be thirty-six], ...is who runs the business...'
The noun phrase antecedent "El hijo mayor/The oldest son" is made relevant by
means of its semantic link with the mention of the speaker's children in the immediately
anterior context.
Fox and Thompson claimed that all noun phrases that contained RCs in their data
were grounded (300). There were instances in my data in which none o f the types of
Grounding explained above occurred. Consider the examples in (103) and (104):
(103) Toda la naturaleza entrega sus dones a los hombres. Son los
hombres [que violan la ley, malinterpretan,...] (Lope Blanch, El Habla de la
Ciudad 118)
■Nature provides its gifts to men. It is men [who violate the law...]'
(104) Enc.-i,Cantan todos los [que van a la iglesia]?
Inf.- Si, cantan todos los [que van a la iglesia]. (Lope Blanch, El Espanol 132)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
64
'Enc.-do all the people [who attend church] sing?
Inf.- Yes, all the people [who attend church] sing'
In (103), the antecedent "los hombres" is a repetition of a noun phrase (in bold)
that was produced in the immediate anterior context. Similarly, in (104), the antecedent of
the RC is a repetition of the immediately anterior context (in bold). In this example, both
the RC and its antecedent are exactly as they were produced in the immediate earlier
context. I considered cases such as the ones in (103) and (104) examples in which
repetitions of noun phrase antecedents of RCs made their grounding unnecessary.
2.2.6. Functional Role of the Relative Clause with respect to its antecedent
I will analyze the correlation o f the 11 TCs with various types of functional roles
that RCs can have with respect to Human and Nonhuman antecedents. After coding all
the data, I have identified four possible types of functions of the RC: Characterize its
antecedent, Identify its antecedent, Give New Information about its antecedent, and
Reaffirm the existence o f its antecedent.
(105) shows a case in which an RC characterizes its antecedent:
(105) Enc.-cNo son daninos los tintes de pelo?
Inf.-No, no. Son sales de amonio cuatemarias. [que son totalmente inofensivas;
oxidates, base de ox...de entre oxido...] (Lope Blanch, El Habla de la Ciudad 39)
Tinc.-Arerit hair dyes harmful?
Inf.-No, no. quaternary ammoniac salts, [that are totally harmless;
oxidant, based on ox...among oxides...]'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
65
The RC "[que son totalmente inofensivas]/[that are totally harmless;...]" gives a
Characterization or description of the referent of the nominal phrase antecedent "sales de
amonio cuatemarias/quatemary amoniac salts."
I use Fox and Thompson's (306-7) definition of Characterization: a
Characterization is typically a predicate that expresses “habitual attributes or properties or
describe [s] features of [its] subjects.” (1990: 307) Characterizations are expressed mostly
by means of intransitive predicates, like those cited by Fox and Thompson (307):
she teaches at the University of Colorado
is a linguist
works on Indonesian
goes to LS A meetings
is an Austronesianist
sleeps late on weekends
likes to dance
etc.
Characterization is one of the two types of function that Fox and Thompson (301-
2) discussed in their investigation of conversational English RCs. Specifically, these
authors referred to the function of an RC of characterizing a New referent of a nominal
phrase. The example given by these authors is shown in (106):
(106) There is a woman in my class [who's a nurse] (301)
In (106), the RC characterizes or describes a referent 'not previously known to the
hearer’ (Fox and Thompson 301).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
66
Another function that RCs perform in the data o f this investigation is the one of
Identifying an antecedent. Fox and Thompson described a second function o f the RC,
according to which, “the relative clause helps to identify a Given Head NP referent,
previously known to the hearer” (302). (107) shows the example given by the authors:
(107) and then the one [that's bigoted], she's married to this guy (Fox and Thompson
302)
Note that the information contained within the RC in (107) also provides a
characteristic feature o f its noun phrase antecedent, that is, the fact that the woman in the
example is a bigoted person.
In order to consider that a RC has the flxnction of identifying its antecedent, I had
two requisites in my investigation. The first requisite is that the referent o f the noun
phrase antecedent o f an RC has to be Evoked (Textually or Situationally Evoked). The
information contained in the RC has to be also Textually or Situationally Evoked. In
order to illustrate these two requisites consider the example in (108):
(108) ...pues que eran unos enamorados, y que la muchacha todos los dias esperaba al
muchacho en su ventana, y una vez el muchacho tuvo que salir de viaje, y salio y
se tardo mucho en regresar. Entonces ella, cuando el llego, ella ya se habia muerto
se supone, <^no? —, puesto que hasta la fecha se ve que el individuo, como a las
doce de la noche, asi cuando dan las doce en el Convento de San Jacinto, se
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
67
aparece por la callecita asi en su caballo, y se va a la ventana [donde ella lo
esperaba] (Lope Blanch, El Habla de la Ciudad 52)
'...they were a couple, and the girl waited for the guy every day at her window,
and once the guy had to go on a trip, and he left and he took too long to come
back. Therefore, when he came back, she had already died, right? — , because to
this days, at approximately twelve at the Convent of San Jacinto, he shows up at
the little street on his horse, and he goes to the window [in which she used to wait
for him]' (Lope Blanch, El Habla de la Ciudad 52)1 8
In the context that precedes the nominal phrase "la ventana/the window," its
referent was already mentioned, along with the information that the girl used to wait for
the guy. Therefore, the information both in the nominal phrase antecedent and within its
RC are retrieved from the anterior context, and both constitute Evoked information for
the hearer.
An identification can also occur when a RC and its antecedent are Situationally
Evoked, as can be observed in (109):
(109 )...£Tiro usie las tortillas \ q u ' estaban en el refrigerador]? (Lope Blanch, El Habla
Popular 414).
'...Did you throw away the tortillas [that were in the refrigerator]?'
1 8 In order to exclude linguistic material that is not relevant for the analysis,
I use the notation'—' in some examples from my data.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6 8
Before the utterance of the example in (109), there was not a previous mention to
las tortillas/the tortillas', but both the antecedent and its RC in (109) were evoked by the
hearer because o f an earlier situation in which she had seen the referent the speaker was
talking about, that is, the tortillas.
A third function that was found in the data o f this investigation is the one in which
the primary function o f the RC to give Brand-new information about its noun phrase
antecedent. The example in (110) illustrates this function:
(110) ...orita estamos peleando un poco, entre los mismos mejicanos. Mejicano vive...ya
se dio cuenta uste aqui, aqui, en esta area de aqui, baja de San Marcos,...En las
lomas, alia viven el americano y uno que otro meiicano [que se sale de aqui],...se
pone mejorcito y se va p'alla. (Lope Blanch, El Espanol 157)
'...nowadays we Mexicans, are fighting amongst each other.
Mexicans live, you already noticed here, they live here, in this lower area o f San
Marcos, ...In the hills, they live there, the Americans and some Mexicans [that
move out of this area], ...their situation improves and they go to settle over there.'
In (110), the RC is giving information that is Brand-new for the hearer and it is, at
the same time, restricting the reference of its antecedent. The RC restricts its antecedent
to a number o f Mexicans who tend to leave the area. It does not identify its antecedent
because both the RC and its antecedent are not Evoked referents. The RC does not have
a characterizing function either, since it does not provide a descriptive feature of its
antecedent.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
69
A fourth type o f function that I found in the data is the one in which the relative
clause has the function o f Reaffirming the existence o f its antecedent, as it is shown in
(111):
(111) el ambiente en que yo me crie, desde muy chico, fiie un
ambiente bastante sano, sobre todo era...pues, limpio, fuera de
envidias-digamos-de las envidias normales [que puede haber] (Lope Blanch, El
Habla de la Ciudad 11-12)
the environment in which I was raised, since I was very young, was a very healthy
environment, with no jealousy-I say-of the normal jealousy [that can exist].'
The information that the RC provides about its antecedent is not Brand-new for
the hearer. This is not a case in which there is a function of Characterization or
Identification. Instead, the function of the RC is to Reaffirm the existence of its
antecedent, that is, to reaffirm that normal jealousy can exist. The information given by
the RC is indeed Inferable for the hearer from the anterior context because of the earlier
mention to the environment in which the speaker grew up. When the speaker mentioned
an environment "with no jealousy," the existence of the concept of "the normal jealousy"
can be inferred. In (111) the use of an existential verb in the RC makes the existence of
the antecedent explicit.
2.2.7. Informational Value of the relative clause with respect to its antecedent
The variable Informational Value of the RC with respect to its antecedent has been
coded separately from the Functional Role of the RC. As I explained earlier, the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
70
informational value o f a RC is always Evoked when the functional role of the RC is to
identify its antecedent. According to this, in the example in (108), in which the functional
role of the RC is the one o f identifying its antecedent ("la ventana/ the window"), the
informational value o f the RC is Evoked.
However, in the case of the function of Characterization, the informational value
of a RC may be Brand-new, Inferable or Evoked.
When the primary function of a RC is to Give New Information about its
antecedent, the informational value of the RC is always Brand-new. In (110) the
informational value o f the RC ("[que sale de aqui]/[that move out of this area]") was
coded as Brand-new. On the other hand, in (111) the informational value of the RC (' [que
puede haber]/[that can exist]’) was coded as Inferable.
In Table 1 ,1 summarize the different functional roles that RCs can have. This
summary also shows the information status of an antecedent and informational value o f a
RC associated with each functional role.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
71
Table 1
Functional roles of a Relative Clause, Information Status of an antecedent, and
Informational Value of a Relative Clause
Functional Role Information Status
of the antecedent
Informational Value
of the Relative
Clause
Characterization Brand-new, Evoked,
or Inferable
Brand-new, Evoked,
or Inferable
Identification Evoked Evoked
Give New Information
about an antecedent
Brand-new, Evoked,
or Inferable
Brand-new
Reaffirm the existence
of an antecedent
Inferable or Evoked
antecedents
Inferable
2.3. Presence/Absence of preposition with a relativizer
I will consider the following possibilities within this variable: Presence/Absence of
a preposition with the relativizer "que/that," Presence/Absence of a preposition with the
relativizer "donde/where," and the cases in which the altemance of Presence/Absence o f a
preposition before a relativizer is not grammatically possible.
Real Academia Espanola refers to the mandatory use o f prepositions when the
relativizer "que" is a circumstantial complement (529), as illustrated in (112):
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
72
(112) Se pasaron quince dias [en que no la vimos] (Real Academia Espanola 529, citing
Cervantes, Quijote, I, 40).
Tifteen days [in which we did not see her] passed.1
Real Academia Espanola also mentions the very common use o f "que," omitting
the preposition when circumstances of time and place are expressed in the antecedent
(529) as in (113):
(113) Hace tres anos [que no lo veo]. (Real Academia Espanola 529)
I t has been three years fin which I have not seen him].'
Real Academia Espanola also refers to the omission of prepositions if the
antecedent has the same preposition (529). In the data of this investigation there are cases
like the one shown in (114), in which the antecedent of an RC does not contain a
preposition, but the speaker omits the preposition in the relativizer:
(114) ...a tal grado que un dia [< p que sail yo de la Escuela], me habian rajado con
una...este navaja la llanta del carro, ... (Lope Blanch, El Habla de la Ciudad 17).
'...to the point in which one day [in which I left the School], they had slashed my
car tire with a blade.'
(115) and (116) shows cases in which there are Presence and Absence of a
preposition before the adverbial relativizer "donde":
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
73
(115) Ahi es-digamos- un seminario [en donde se recibe la formacion general para el
sacerdocio, particularmente en el aspecto espiritual,...] (Lope Blanch, El Habla de
la Ciudad 56).
I t is -I sav- a seminary [in which the general instruction for priesthood is received,
particularly in the spiritual aspect,...]'
(116) Ayudamos a todos los ninos que estan hospitalizados y a los de consulta externa
del hospital, [donde los ninos son pobres, desde luego] (Lope Blanch, El Habla de
la Ciudad 91).
W e help all the children that are hospitalized and the ones o f external
consultation of the hospital, [where the children are poor, of course]1
Finally, there are cases that I coded as "does not apply" because they are instances
in which the altemance between Presence/Absence of a preposition before a relativizer is
not possible without interfering with the content of what is being expressed within the RC.
One of these cases is illustrated in (117):
(117) ...nos enteramos de que el trabaio [que hacian] era muy...bonito,... (Lope Blanch,
El Habla de la Ciudad 86).
'...we found out that the work [that they did] was very...nice,...1
In (117) there is no possible alternation between the Presence and the Absence o f a
preposition before the relativizer without changing the semantic content of the idea
expressed within the RC. Consider (118), which is (117) with the addition of a
preposition with the relativizer:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
74
(118) ...nos enteramos de que el trabaio [de/por/a/en que hacian] era muy...bonito,...'
'...we found out that the work [of/for/to/in that they did] was very...nice,...’
The relativized noun phrase in (117) has the grammatical role o f direct object, but
in (118) each o f the added prepositions alters the grammatical role within the RC and the
meaning o f what is expressed by the RC.
The variable Presence/Absence of Preposition will be analyzed across different
socioeconomic levels and different educational levels o f the speakers in order to
investigate a possible correlation of lower socioeconomic level and lower level of
education with the use of the variant that grammars consider non standard: the absence of
a preposition with a relativizer when the alternation is possible.
2.4. Register
In this study, the terms "register" and "register variation" will be used, following
Finegan and Biber*s definition, "as cover terms for the full range of language varieties
associated with differences in communicative situation (including mode and purpose)."
(316).
The variable Presence/Absence of a preposition with a relativizer will be analyzed
in different registers, which will be summarized as: One to one interview, Group
interview, Monologue, and Conferences.
The variant Presence of a preposition will be analyzed as a standard variant that is
expected to exhibit high frequencies of use in types o f registers in which the use of
standard variants is typical, such as Conferences.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
75
2.5. Social variables
The social variables that I analyzed in this study are: socioeconomic level, level of
education, and the variable that I labeled Monolingual/Bilingual speaker (see Appendix 3
for the distribution o f the speakers within each of these social variables). In Silva-
Corvalan's words:
the physical context in which communication takes place, the relationship
among the speakers and both their 'attributed' social characteristics
(generational group, sex, ethnicity, cast, etc) and 'acquired' (educational
level, socioeconomic level, etc.), have been proved to be systematically
reflected in differentiated linguistic behavior. This covariation between
linguistic and social phenomena is what defines the 'sociolinguistic variable.
(Sociolineuistica 68)1 9
I will investigate the influence of Level of education and Socioeconomic level on
the use o f Restrictive and Nonrestrictive RCs and on Presence/Absence of a preposition.
In this investigation I classify the speakers using the information available with every
transcription. The criteria used to classify the speakers in either lower or middle class are
the type of occupation or profession and types of activities that the speaker describes,
along with the types of social circles he interacts with in his/her daily life.
As I have explained in Chapter 1, several studies have found that relativization is a
grammatical feature that correlates with factors such as economic status and level of
education. These studies justify my interest in investigating, among the three groups of
speakers in my data, possible differences in the production of Restrictives and
l9 My translation o f Silva-Corvalan fSociolinguistica 68).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
76
Nonrestrictive RCs in discourse, and in the variable Presence/Absence of a preposition
with a relativizer.
The question I hope to answer is: do the speakers who belong to the middle class
and the speakers with higher levels o f education tend to produce a higher percentage of
Nonrestrictive RCs than the speakers who belong to the lower class? Another question in
this investigation regarding Socioeconomic level and Level of education is: do lower
levels o f education and a lower socioeconomic level tend to favor the absence of a
preposition before a relativizer?
The bilingual speakers in this study are comparable to the monolingual lower class
speakers from Mexico City, who have had little formal education in Spanish. These two
groups of speakers will be analyzed in their use of a preposition with a relativizer.
Because these two groups share a lower level of education in Spanish, it is expected that
they will behave similarly. This will be examined in Chapter 4.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3
RESULTS OF THE ANALYSIS OF LINGUISTIC VARIABLES
77
3.1. Introduction
In order to analyze grammatical variation, it is necessary to pay attention to
semantic, pragmatic, and functional factors. The interaction o f these factors is what
underlies the grammatical variation observed in relativization.
In this chapter I examine the correlations among semantic/pragmatic factors and
the 11 TCs with more frequent occurrences in my data.
In order to analyze the patterns o f variation of the 11 TCs, I consider the syntactic
roles o f both the Nominal Phrase antecedents (Head NP roles, in Fox and Thompson's
terminology (298)) and the coreferent nominal phrase within the RC (NPRel in Fox and
Thompson (298)).
The combination of grammatical roles is studied in this chapter, taking into
account that in the interaction between speaker and hearer there is a sequence in the
production of a RC in which the hearer first hears the nominal phrase antecedent
(antecedent in what follows) and then he/she hears the RC. This is Fox and Thompson's
perspective, which they call “a real-time processing perspective.” (303) For these reasons,
the semantic/pragmatic characteristics o f the nominal phrase antecedent are the first to be
analyzed in order to determine their relevance in the speakers' selection o f the type of RC
that they use in each case.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
78
3.2. Other Studies
Few studies consider semantic/pragmatic factors o f the interaction as relevant in
the syntactic variation in RCs. In English, Lehman studied the stress factor in terms of the
function and position o f RCs (Fox and Thompson 297).
The direct antecedent o f my study is the investigation by Fox and Thompson,
which analyzed patterns of variation o f RCs that are clearly influenced by Information
Flow factors such as Humanness, Definiteness, Information Status of the antecedent of the
RC, Grounding and Functional Role o f the RC with respect to its antecedent. The authors
analyzed the TCs Subject-Direct Object (s-do) and Direct Object-Direct Object (do-do)
for Nonhuman antecedents, and Existential-Subject (ex-s) for Human antecedents. In this
chapter I show that the patterns o f distribution o f the 11 TCs ( Restrictive and
Nonrestrictive RCs) correlate with factors such as Humanness, Information Status,
Grounding and Functional Role o f the RC.
3.3. Analysis
3.3.1. Humanness
My first hypothesis is that the factor Humanness determines different patterns of
relativization for Human and Nonhuman antecedents o f RCs. Rosenbaum (cited by
Macaulay 64) observed the tendency for RCs with Human antecedents to have the role of
Subject and for RCs with Nonhuman antecedents to have the role of Direct Object.
Macaulay confirmed this observation in his study of RCs in Ayr, Scotland (Macaulay 64).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
79
As noted by Silva-Corvalan, in Spanish the tendency is for Subjects to be human,
definite and preverbal ("On the Interaction" 119). Direct Objects, however, are often
indefinite, inanimate and postverbal. Given these general tendencies observed in Spanish,
it is expected that the syntactic roles of both an RC and its antecedent will be influenced
by the Humanness of the noun phrase antecedent.
To test this hypothesis, I establish a correlation between the factor Humanness and
the 11 TCs in Table 2:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TABLE 2
80
TYPE OF COMBINATION BY HUMANNESS
Type o f
Combination
Humanness Total
N
Human
%
Nonhuman
N %
N %
cc-s 10 1% 106 6% 116 4%
cc-cc 3 0% 301 17% 304 12%
OP-S 144 17% 143 8% 287 11%
OP-DO 16 2% 154 9% 170 6%
PN-S 154 19% 183 10% 337 13%
PN-DO 14 2% 158 9% 172 7%
DO-S 141 17% 197 11% 338 13%
DO-DO 17 2% 203 11% 220 8%
s -s 218 26% 161 9% 379 14%
S-DO 17 2% 133 7% 150 6%
EX-S 99 12% 77 4% 176 7%
TOTAL 832 100% 1816 100% 2649 100%
p<.000
Table 2 shows that for Human antecedents, every syntactic role of the antecedent
of a RC strongly favors Subject relativization over Object relativization. I use the label
"Subject relative" to refer to a RC in which the relativized position has the role of Subject.
The label "Object relative" is used to refer to those RCs in which the relativized position
has the role of Direct Object (Fox and Thompson 298). Similarly, in order to refer to the
RC in the TC 'cc-cc1 ,1 will use the term "Circumstantial Complement relative" (CC
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8 1
relative in some tables in this chapter). Moreover, the most common grammatical role o f
Human antecedents is Subject, as (119) shows:
(119) Human Antecedents in Table 2
Subject antecedents (for TCs s-s and s-do): 28%
Predicate Nominal antecedents (for TCs pn-s and pn-do): 21%
Direct Object antecedents (for TCs do-s and do-do): 19%
Object of Preposition Antecedent (for TCs op-s and op-do): 19%
Existential Antecedent (for TC ex-s): 12%
Circumstantial Complement Antecedent (for TCs cc-s and cc-cc): 1%
Table 2 reveals that the TC most favored by Human antecedents is 's-s,' as in
( 120):
(120) ...el padre [que me dirigia en la clase] dijo:... (Lope Blanch, El Habla de la Ciudad
130)
'...the priest [that guided me in the class] said:...'
The example in (120) shows a Human referent antecedent with the role o f subject,
which is relativized in the role of Subject.
In summary, the observations made on Table 1 reveal that Human antecedents of
RCs and their relativized coreferents follow the general tendency in Spanish to be
Subjects, as observed by Silva-Corvalan for main clauses ("On the Interaction" 119).
For Nonhuman antecedents, Table 2 shows that Subject antecedents are one of the
two least frequent roles in the antecedent of RCs, as I show in (121):
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
82
(121) Nonhuman antecedents in Table 2
Circumstantial Complement Antecedent (for TCs cc-s and cc-cc): 23%
Direct Object Antecedent (for TCs do-s and do-do): 22%
Predicate Nominal Antecedent (for TCs pn-s and pn-do): 19%
Object o f Preposition Antecedent (for TCs op-s and op-do): 17%
Subject Antecedent (for TC s-s and s-do) : 16%
Existential Antecedent (for TC ex-s): 4%
Circumstantial Complement and Direct Object are the most frequent syntactic roles
for Nonhuman antecedents of RCs. The high frequency of Circumstantial Complement
antecedents with Nonhuman referents is expected since Circumstantial Complements refer
to places, time, mode, instrument, cause or means (Real Academia Espanola 375), which
tend to favor Nonhuman referents.
It is also not surprising that Direct Object antecedents are almost as frequent as
Circumstantial Complements as antecedents of RCs with Nonhuman antecedents. This
observation only confirms the general tendency for Nonhuman entities to be coded as
Direct Objects in Spanish (Silva-Corvalan, "On the Interaction" 119). Object relatives do
not outnumber Subject relatives when antecedents are Nonhuman. For each syntactic role
in the antecedent, the percentages for Subject and Object relatives are very close (36% of
Object relatives and 38% of Subject relatives), as (122) (A) shows:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
83
(122) Subject relativization and Objet relativization with
Nonhuman and Human antecedents in Table 2
(A) Nonhuman antecedents (B) Human antecedents
Object of Preposition-Subject 8% Object o f Preposition-Subject 17%
Object of Preposition-Direct
Object 9%
Object o f Preposition-Direct
Object 2%
Predicate Nominal-Subject 10% Predicate Nominal-Subject 19%
Predicate Nominal-Direct
Object 9%
Predicate Nominal-Direct
Object 2%
Direct Object-Subject 11% Direct Object-Subject 17%
Direct Object-Direct Object 11% Direct Object-Direct Object 2%
Subject-Subject 9% Subject-Subject 26%
Subject-Direct Object 7% Subject-Direct Object 2%
However, it is important to notice that when Humans and Nonhumans are
compared in Table 1, with Nonhuman antecedents the percentage of Object relatives
increases with respect to the percentage of Subject relatives. If (122)(A) and (122) (B)
are compared, the distance between percentages for Subject and Object relatives is bigger
for Human antecedents, as shown in (122) (B). On the other hand, the percentages for
Subject relatives and Object relatives are very close for each TC with Nonhuman
antecedents, as shown in (122) (A).
In sum, (122) (A) and (122) (B) reveal that while Nonhuman antecedents may be
relativized as Subject or Direct Object relatives, Human antecedents are rarely relativized
as Direct Objects.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
84
With Nonhuman Circumstantial Complement antecedents, there is a clear tendency
for Circumstantial Complement relatives over Subject relatives (17% and 6%
respectively). The TC 'cc-cc' with a Nonhuman antecedent is shown in (123):
(123) La aventura estaba hecha, desde el momento mismo [en que quemamos nuestras
naves...] (Lope Blanch, El Habla de la Ciudad 102).
'The adventure was done, since the very first moment [in which we wrapped up all
our business here ....]'
In example (123), the antecedent of the RC expresses a circumstance of time,
which has the same role within the RC. The Nonhuman feature that a Circumstantial
Complement antecedent typically has determines that usually the role within the RC will
be replicated as a Circumstantial Complement. Notice that in (123) the referent o f the
Circumstantial Complement antecedent has the same syntactic role within the RC, that is,
it expresses the same circumstance of time within the RC.
In summary, the data in Table 2 strongly supports the influence o f the factor
Humanness on the TCs of RCs. Humans tend to be coded as Subjects both in the
antecedents o f RCs and in the relativized positions.
Nonhuman antecedents favor the role Direct Object over Subject, but Direct
Object antecedents are almost as common as Circumstantial Complement antecedents.
The role of Circumstantial Complement (23%) was used by the speakers almost as
frequently as the role of Direct Object (22%) with Nonhuman antecedents.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
85
3.3.2. Grounding and Humanness
My second hypothesis is that speakers select different types of Grounding for
Human and Nonhuman antecedents o f the 11 TCs that I study. Fox and Thompson (303)
showed the tendency for Object relatives to perform Anchoring when the antecedent is a
Nonhuman Subject.2 0 In my investigation, I found that Anchoring is associated with the
TC 's-do' with both Human and Nonhuman antecedents, which allows me to extend Fox
and Thompson's observation to Human and Nonhuman antecedents, at least for Spanish.
Fox and Thompson also noted the statistical association of the TC 'do-s' with Main-Clause
grounding of Nonhuman antecedents (304-05). Usually the Subject in the main clause
(which contains the RC) grounds the antecedent o f the RC, before the utterance o f the
RC. For this reason there is no need for the RC to provide the grounding. My data
showed that the TC 'do-s' also favored Main-Clause grounding, but this pattern is the
same for Human and Nonhuman antecedents.
Fox and Thompson noticed that the Existential antecedents in their data were
Indefinite and Human (309). Human referents c < tend to be grounded by being related to
their own activities, that is, to earlier predicates, as with a locative or by proposition-
linking.” (309) They concluded that the discoursive need o f relating a Human referent to
2 0 Fox and Thompson only focused on Nonhuman antecedents in their
analysis o f antecedents with the role o f Subject and Direct Object. For antecedents
with the role of Existential, the authors focused on Human antecedents (the
Existential antecedents in their study were Human).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
86
its own activities is what tends to trigger Subject relatives rather than Object relatives (309).
In my study the TC 'ex-s' (Human and Nonhuman antecedents) tends to strongly
favor Proposition-Linking grounding. This suggests that Subject relatives are more
appropriate than Object relatives when there is Proposition-Linking grounding because
with this type of Grounding there is no need for Human referents (that are different from
the antecedents that need Grounding) to provide the Grounding. Within Subject relatives,
typically the referent that has the role of Subject is coreferent with the antecedent of the
RC. Consider, for instance, the example in (124), which shows a Subject relative in a case
of Proposition-Linking grounding:
(124) Ayer estuve buscando coches. Habia un coche [ que tenia cuatro puertas], pero yo
buscaba algo mas pequeno.
'I was looking for a car yesterday. There was a car [that had four doors], but I
was looking for something smaller1 .
In (124), the speaker had said that he/she was looking for a car. Therefore, when
the noun phrase "un coche/ a car" is mentioned, its referent is linked to the earlier
proposition "estuve buscando coches/ I was looking for a car." Notice that in the
Existential construction "habia un coche/there was a car" the noun phrase antecedent
follows the Existential verb "habia." In (124), and generally in an Existential phrase, there
are no additional referents with which an antecedent of an RC can be related in order to be
grounded before the utterance of the RC. Typically an Existential structure consists o f an
Existential (generally "haber" in my data) that is followed by a noun phrase, as in (124).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
87
The Subject relative in (124), "[que tiene cuatro puertas]/[that has four doors]" has
a subject that is coreferent with the antecedent of the RC. Notice that within the Subject
relative in (124), there cannot be another Subject pronoun (other than the coreferent of
the antecedent) that may ground that antecedent.
Unlike Subject relatives, Object relatives have a type o f structure that allows them
to provide Grounding. The example in (125) illustrates this:
(125) Una clase [que (yo) tomo este semestre] es muy interesante.
'A class [that I am taking this semester] is very interesting'.
In the TC in (125), the RC grounds (by Anchoring) its antecedent, by means of the
implicit Evoked referent of "yo/I." Assuming that there was no Proposition-Linking
grounding, the referent of "una clase/a class" is not grounded until the utterance of the RC
"[que tomo este semestre]/[that take this semester]." For this reason, the implicit subject
"yo/I" within the Object relative anchors its antecedent. Indeed, Fox and Thompson
noticed that Object relatives were better suited than Subject relatives for the task of
Anchoring Nonhuman Subject antecedents (303-04).
In Table 3 ,1 establish a correlation between the three types of grounding and the
TCs with Human antecedents. In Table 4 ,1 establish the same correlation but with
Nonhuman antecedents. A total of 80 RCs were not included in these correlations
because they occurred with no Grounding.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
TABLE 3
GROUNDING BY TC BY HUMAN ANTECEDENTS
TYPE OF TYPE OF COMBINATION
GROUNDING
cc-s OP-S OP-DO PN-S DO-S
N % N % N % N % N %
Anchoring 2 20% 5 4% 4 27% 3 2% 7 5%
Main-Clause Grounding 5 50% 66 49% 7 47% 4 3% 77 56%
Proposition-Linking
Grounding
3 30% 65 48% 4 27% 141 9 5 % 54 39%
Total 10 100% 136 100% 15 100% 148 100% 138 100%
00
00
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
TABLE 3 (CONTINUED)
Type o f Grounding TYPE OF COMBINATION
DO-DO S-S S-DO EX-S TOTAL
N % N % N % N % N %
Anchoring 3 18% 29 14% 11 69% 11 12% 75 10%
Main-Clause Grounding 5 29% 28 14% 2 13% 6 7% 200 26%
Proposition-Linking
Grounding
9 53% 149 72% 3 19% 74 81% 502 65%
Total 17 100% 206 100% 16 100% 91 100% 777 100%
p<.000
Note; 8 cells (29,6%) have expected count less than 5,
00
vo
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
TABLE 4
GROUNDING BY TC OF NONHUMAN ANTECEDENTS
TYPE OF
GROUNDI
NG
TYPE OF COMBINATION
CC-S cc-cc OP-S OP-DO PN-S PN-DO
N % N % N % N % N % N %
Anchoring 12 12% 113 38% 12 9% 44 29% 2 1% 17 11%
Main-
Clause
Grounding
45 45% 109 37% 63 45% 42 27% 3 2% 3 2%
Proposition-
Linking
Grounding
44 44% 73 25% 65 46% 68 44% 175 97% 135 87%
Total 101 100% 295 100% 140 100% 154 100% 180 100% 155 100%
VO
o
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
TABLE 4 (Continued)
TYPE OF TYPE OF COMBINATION
GROUNDING
DO-S DO-DO S-S S-DO EX-S TOTAL
N % N % N % N % N % N %
Anchoring 4 2% 20 10% 30 20% 69 52% 9 12% 332 19%
Main-Clause
Grounding
112 57% 67 34% 18 12% 20 15% 7 10% 489 28%
Proposition-
Linking
Grounding
79 41% 112 56% 101 68% 44 33% 58 78% 954 54%
Total 195 100% 199 100% 149 100% 133 100% 74 100% 1775 100%
pc.000
92
A first pattern that can be observed in Table 3 is that the TCs that have Object
relatives have higher percentages for Anchoring (as compared to other types o f
grounding) than the TCs that have Subject relatives. For every grammatical role in a noun
phrase antecedent in Table 3, the percentage of Object relatives is higher than the
percentage of Subject relatives when there is Anchoring. The association o f Object
relatives rather than Subject relatives with Anchoring was noticed, as I have already
explained, by Fox and Thompson (302-03) with Nonhuman antecedents with the role of
Subject ('s-do'). Because the antecedent with the role o f Subject has an early position in
the sentence, its referent usually has not been grounded until the utterance o f the RC. The
RC grounds the antecedent by Anchoring. In the present study, Table 3 reveals that for
antecedents with the roles of Object o f Preposition, Direct Object, and Subject, Object
relatives are preferred over Subject relatives for Anchoring Human antecedents. As noted
by Fox and Thompson (300), a referent is anchored by an Evoked referent within the
Object relative (Given referent in Fox and Thompson's terminology). The Anchoring
occurs from the RC, by an Evoked subject pronoun within the RC that is different from
the referent that needs grounding, as in (128)2 1 :
(128) La muier [que (yo) invite] no vino.
'The woman [that I invited] did not come over1 .
Assuming that the antecedent "la mujer/the woman" has not been grounded by the
time the RC is uttered, the Evoked implicit subject pronoun "yo/I" within the RC anchors
zlExamples (126) and (127) have been intentionally omitted.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
93
the antecedent "la mujer/ the woman." This relation between an Evoked anchor and the
antecedent is what makes the referent o f that antecedent relevant in the discourse. As was
noted by Fox and Thompson, Object relatives typically contain a pronominal reference
(Human) to anchor the antecedent, and a Nonhuman coreferent of the antecedent within
the RC (304). This makes Object relatives “the most common mechanism for anchoring
referents.” (304)
A second pattern in Table 3 is that when comparing the pairs o f TCs 'do-do' / 's-do'
and 'do-s' /Vs,' the TCs in which the antecedent has the role of Direct Object rather than
Subject tend to be associated with Main-Clause grounding.
A third pattern in Table 3 is that the TCs 'pn-s,' 's-s,1 and 'ex-s' strongly favor
Proposition-Linking grounding. As I will explain later in this section, the grammatical
roles in the antecedent of the RC and within the RC in the TCs 'pn-s,' 's-s,' and 'ex-s'
interact with the need to ground a referent without relating such referent to other
referents, as with Anchoring and Main-Clause grounding.
Table 4 (Nonhuman antecedents) exhibits the same patterns observed in Table 3
for Human antecedent, that is, for the syntactic roles of Object of Preposition, Direct
Object, and Subject in the antecedent, Object relatives are more frequent than Subject
relatives with Anchoring. When comparing the pairs o f Tcs 'do-s'/ 's-s' and 'do-do' / 's-do,'
the TCs with antecedents with the role o f Direct Object tend to be more frequent with
Main-Clause grounding. The most obvious pattern is that the Tcs 'pn-s,' 'pn-do,' 's-s,' and
'ex-s' strongly favor Proposition-Linking grounding.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
94
Table 3 (Human antecedents) does not include the TC 'cc-cc' because of the
limited amount o f cases of this TC with Human antecedents (3 cases), which caused an
empty cell in an initial correlation. As (129) shows, in some circumstances, a verb may
have a Human referent:
(129) ...voy con las madres salesianas. [donde se estaban educando...] (Lope Blanch, El
Habla de la Ciudad 140)
'I went to the nuns o f the Salesian order, [where they were being educated]
In (129), the speaker indicated a place she visited, which is a school managed by
nuns of the Salesian order in Mexico, where her daughters were being educated.
Although the antecedent of the RC indicates a place, it was expressed in terms of the
people who managed it: the nuns of the Salesian order.
Table 3 does not include the TC 'pn-do' either. The reason is that the 14 cases o f
this TC with Human antecedents occurred with one type of grounding, which was
Proposition-Linking grounding. The fact that there were no cases of Anchoring or Main-
Clause grounding caused the existence of empty cells in the initial correlation. As I will
explain later, the grammatical role of Predicate Nominal is strongly associated with
Proposition-Linking grounding.
In Table 4, the percentages for Anchoring and Main-Clause grounding for the TC
'cc-cc' are very close (38% and 37% respectively). A lower percentage (25%) of the cases
of'cc-cc' was grounded by Proposition-Linking grounding. The fact that Humans tend to
ground Nonhuman referents explains the preference for Anchoring and Main-Clause
grounding of Nonhuman antecedents, as (130) and (131) show:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
95
(130) CC-CC
En una ocasion va...por el ano de cuarenta v...cuatro...ren que (yo)hice un viaje a
los Estados Unidos...] (Lope Blanch, El Habla de la Ciudad 101)
'In one occasion already around the year forty for... [in which I made a trip to the
United States...]'
(131) CC-CC
Enc.-i, Y tu te acuerdas de tu papa o no?
Inf.-No, no lo conoci. Pues (el) se murio el mismo ano [que yo naci.] (Lope
Blanch, El Habla de la Ciudad 143)
'Erie.- And do you remember your dad or not?
Inf.- No, I did not meet him. Because he died the same year [in which I was bom]
In (130), Anchoring occurs by means of the referent of the implicit pronoun "yo/I"
within RC. In (131), Grounding occurs earlier by means of Main-Clause grounding. The
evoked (implicit) subject pronoun "el/he" (which refers to the speaker’s father) is the
element that grounds the antecedent "el mismo ano/the same year" in the main clause.
In (132) I show the tendency observed for Human antecedents to be anchored by
Object relatives rather than by Subject relatives:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
96
(132) Anchoring with Human antecedents in Table 3
Percentage of Anchoring as compared to other types of grounding
op-s 4% do-s 5% s-s 14%
op-do 27% do-do 18% s-do 69%
In (132) I did not contrast the TCs 'pn-s1 and 'pn-do,' because, as I have explained
above, all the cases of'pn-do' had Proposition-Linking grounding. (132) shows that
Anchoring of Human referents favors Object relatives over Subject relatives.
As I indicated earlier in this chapter, Table 4 reveals that with Nonhuman
antecedents there is also a tendency for Object relatives rather than Subject relatives to
perform Anchoring, as is shown in (133):
(133) Anchoring with Nonhuman antecedents in Table 4
Percentage of Anchoring as compared to other types of grounding
op-s 9% pn-s 1%
op-do 29% pn-do 11%
do-s 2% s-s 20%
do-do 10% s-do 52%
(133) shows that Nonhuman antecedents also favor Object relatives when there is
grounding by Anchoring. Fox and Thompson (302-303) also observed this tendency for
the TC 's-do' (Nonhuman antecedents), as I mentioned above. On the other hand, their
observation that Nonhuman Subject antecedents tended to occur with Object relatives
rather than with Subject relatives in their data is not replicated in the Spanish data in this
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
97
investigation. As Table 2 shows, 9% of the 11 TCs of my study were 's-s' and 7% were's-
do.' Thus, it cannot be claimed that either Subject relatives or Object relatives are strongly
favored by Nonhuman Subject antecedents.
For Nonhuman Subject antecedents, Fox and Thompson found that 77% of the RC
were Object relatives and 13% were Subject relatives (302). These authors claimed that
the preponderance of Object relatives with Nonhuman Subject antecedents was
determined by the speakers' need to ground Nonhuman antecedents by Anchoring (303).
The results in this chapter reveal that this cannot be claimed for Spanish.
Tables 3 and 4 reveal that for both Human and Nonhuman referents, the speakers
prefer to use Object relatives rather than Subject relatives for Anchoring. Consider the
examples in (134) and (135), which show a case of Anchoring with an Object relative and
with a Subject relative, respectively:
(134) S-DO
Un cochero [que (nosotros) nos conseguimos en Sevilla]... Nos llevo a verla...
(Lope Blanch, El Habla de la Ciudad 177).
'A carriage driver [that we got in Seville]...took us to see her...'
(135) S-S
Y entonces el padre [que me dirigia en la clase] dijo: ... (Lope Blanch, El Habla de
la Ciudad 130).
'And then the father [that guided me in the class] said:...'
In (134), the implicit Evoked subject pronoun 'nosotros/we' is the element that
anchors the antecedent "un cochero/a carriage driver." In (135), the element performing
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
98
the anchoring within the RC is the pronoun "me/me." In both (134) and (135) there was
an evoked pronoun within the RC (that referred to the speaker) functioning as the element
that achieves the anchoring of the antecedent.
The fact that cases like (134) are more common than cases like (135) with
Anchoring indicates that Anchoring is favored by structures in which the element
performing the Anchoring from the RC is an evoked Human subject pronoun such as the
implicit pronoun "nosotros/we" in (134).
When explaining the preponderance of the TC 's-do' over 's-s' with Nonhuman
antecedents, Fox and Thompson (304) noticed that anchoring is almost always performed
by pronouns, because “pronouns are the primary way to refer to Given referents (Fox
1987).” The pronouns that do the Anchoring generally refer to Humans for two reasons:
the first is that “we humans generally talk about other humans.” (Fox and Thompson 304)
Appendix 4 shows that when there is Anchoring and Main-Clause grounding of
Nonhuman antecedents, there is a very clear tendency for Humans to perform grounding
by Anchoring or grounding by Main-Clause grounding. This same tendency for Humans
to perform the Anchoring from within the RC and to perform Main-Clause grounding is
observed with Human antecedents, but the correlation did not yield significant results.
The second reason for why the pronouns that do the Anchoring are generally
Human is that the Subject antecedents that Fox and Thompson analyzed were Nonhuman.
Fox and Thompson (304) claimed that in conversation, "the major way that nonhuman
referents are made relevant is in terms o f the humans who own them, use them, and
manipulate them (see, e.g., Du Bois 1980: 269-70)." Table 4 (Nonhuman antecedents)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
99
shows that 54% of the cases were grounded by Proposition-Linking, and 47% were
grounded by Anchoring and Main-Clause grounding (19% and 28% respectively).
Because Proposition-Linking grounding does not consist of Humans making Nonhuman
referents relevant, it cannot be claimed that in Spanish Nonhuman referents are mostly
grounded by their relationship with Humans who own them or manipulate them, as
observed by Fox and Thompson (1990) for English.
The Anchoring of a Nonhuman antecedent by an RC that contains an evoked
Human Subject pronoun can be observed in (136):
(136) S-DO
Nada mas el unico problema [que (nosotros) tuvimos] file lo de las bicicletas
(Lope Blanch, El Habla Popular 320)
'The only problem [that we had] was the thing with the bicycles'
In (136) it is clear that when the referent of the implicit Subject pronoun
"nosotros/we" anchors the antecedent "el unico problema/the only problem," a relationship
of ownership is being established between the antecedent and the Subject pronoun that
anchors it. However, as (134) showed, Human pronominal (evoked) subjects also tend to
establish relationships of "manipulation" with Human referents.
Table 4 reveals that Object relatives are not the only type of RC that tends to favor
Anchoring. If the TCs 'cc-s' and 'cc-cc' are compared in Table 4, it can be observed that
the TC 'cc-cc' has a much higher percentage of Anchoring (38%, as compared to the
other types of grounding) than the TC 'cc-s' (12%, as compared to the other types of
grounding). This tendency is due to the fact that Circumstantial Complement relatives,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 0 0
like Object relatives, typically have a Human pronominal subject that can anchor a
Nonhuman antecedent, as (137) shows:
(137) CC-CC
En una ocasion va...por el ano de cuarenta v...cuatro...ren que (yo)hice un viaje a
los Estados Unidos...] (Lope Blanch, El Habla de la Ciudad 101)
In one occasion already around the year forty four... [in which I made a trip to the
United States...]'
In (137), Anchoring occurs by means o f the referent of the implicit pronoun "yo/I"
within the RC. Typically Circumstantial Complement relatives contain Human subject
pronouns that relate to Nonhuman referents within the RC.
Tables 3 and 4 show a pattern that consists of a preference for the grammatical
role o f Direct Object over the role of Subject for Human and Nonhuman antecedents of
RCs when there is Main-Clause grounding. This can be observed in (138) and (139):
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 0 1
(138) Main-Clause grounding with Human antecedents in Table 3
Main-Clause grounding as compared to other types of grounding
do-do 29% do-s 56%
s-do 13% s-s 14%
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 0 2
(139) Main-Clause grounding with Nonhuman antecedents in
Table 4
Main-Clause grounding as compared to other types of grounding
do-do 34% do-s 57%
s-do 15% s-s 12%
(138) and (139) reveal that if the TCs 'do-do' and 's-do' are compared, the TC 'do
do' exhibits a higher percentage of Main-Clause grounding. Similarly, when the TCs 'do-s'
and 's-s' are compared, the TC 'do-s' exhibited a higher percentage o f Main-Clause
grounding.
(140) and (141) show a case of Main-Clause grounding with the TCs 'do-s' and's-
do' respectively:
(140) DO-S
Entonces, (el) encontro una muchacha [que era huerfana tambien de papa y
mama.] (Lope Blanch, El Habla Popular 22)
'Then, he found a girl [that was also an orphan.]'
(141) S-DO
...y mi mama- [que usted la conocio]- habia estudiado en Morelia... (Lope Blanch,
El Habla de la Ciudad 198).
'...and my mother-[whom you met]-had studied in Morelia...'
In (140), the antecedent 'una muchacha/a girl' is the direct object in the main
clause, and gets grounding by the implicit (Evoked) Subject pronoun "el/he." In (141) the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
103
antecedent "mi mama/my mother" does not get Main-Clause grounding by an Evoked
Subject pronoun like in (140). The element that grounds (by Main-Clause grounding) the
antecedent in (141) is the referent o f the possessive pronoun, the speaker, who is talking
about her mother.
The basic difference between the grounding of the antecedent in (140) and (141) is
that in (140) a subject pronoun is the element that does the Grounding, whereas in (141),
a possessive pronoun (relating the speaker to the referent o f her mother) grounds the
antecedent before the utterance o f the RC.
Given the tendency for Subjects to be preverbal and for Direct Objects to be
postverbal in Spanish (Silva-Corvalan, "On the Interaction" 119), typically a Subject (a
Subject pronoun in (140)) will be the element within the main clause that will provide the
grounding to a referent that needs to be grounded. The tendency for antecedents to be
Direct Objects when there is Main-Clause grounding may be influenced by the typical
word order in Spanish. Because o f this typical word order in Spanish, by the time o f the
utterance o f an antecedent with the role o f Direct Object, typically a Subject has already
been produced before the verb. This typical structure is appropriate for a Subject referent
to ground a referent o f a Direct Object that needs grounding.
A third pattern that I observed in Tables 3 and 4 is that the TCs 'pn-s,1 'pn-do,' 'ex-
s,' and 's-s' tend to strongly favor Proposition-Linking grounding. This occurs with
Human and Nonhuman antecedents, as in (142) and (143):
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
104
(142) TCs that favor Proposition-Linking grounding with Human antecedents in
Table 3
Proposition-Linking as compared to other types of grounding
pn-s 95% ex-s 81% s-s 72%
(143) TCs that favor Proposition-Linking grounding among Nonhuman
antecedents in Table 4
Proposition-Linking as compared to other types of grounding
pn-s 97% pn-do 87%
ex-s 78% s-s 68%
(142) does not contain the TC 'pn-do' because, as I have explained above, all the
cases (14 cases) occurred with Proposition-Linking, causing empty cells in the correlation.
However, it is interesting to note this because it indicates a tendency for the TC 'pn-do' to
occur with Proposition-Linking grounding.
(142) and (143) reveal that the grammatical role of predicate nominal in the
antecedent strongly favors Proposition-Linking grounding. (144) and (145) illustrate
cases of Proposition-Linking with the TCs 'pn-s' and 'pn-do' respectively:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
105
(144) PN-S
... este profesore ...doctore...Gascon, que escribio la historia de Nuevo
Mejico...No se si habra leido el libro...Es un libro en espaiiol muv bueno. [que
salio nuevamente aquL] (Lope Blanch, El Espanol 188).
'...this professor...doctor...Gascon, who wrote the history ofNew Mexico...I don't
know whether you have read the book or not...It is a very good book in Spanish.
[which was published here again.]'
(145) PN-DO
Enc.-...pero sus diplomas £como es...
It it
Enc.-^Cual es el primero?...
Inf -Probablemente, extranjero; probablemente fixe un diploma [que me
concedieron en el extranjero en la Universidad Central de Madrid...] (Lope Blanch,
El Habla de la Ciudad 105)
'£«c.-...but your diplomas , how is...?
Which is the first one?...
Inf. -Probably, abroad; it probably was a diploma [that they gave me abroad in the
Universidad Central de Madrid...]
In (144), the antecedent and its RC are grounded by their link to the preceding
context (marked in bold) in which the speaker was talking about a book that he had read,
whose topic is the history ofN ew Mexico. In (145), the antecedent 'un diploma/a
diploma' and its RC are grounded by their link to the immediately anterior context in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
106
which the speaker and the interviewer were referring to the diplomas given to the speaker
during his professional career.
(144) is a typical structure o f Predicate Nominal, which usually does not contain
other referents (different from the antecedent) that can be related to the antecedent. In
(144), there is an equational structure in which the implicit Subject (third person singular)
and the predicate nominal refer to the same entity. The Subject relative "[que salio
nuevamente aqui]" continues referring to the antecedent in terms of itself, without relating
it to other referents within the RC.
The Predicate Nominal structure in (144) does not contain a referent different from
itseif that can ground it before the utterance o f the RC. The RC in (145) is a Subject
relative, which contains a subject that is coreferent with its antecedent ("un libro en
espanol muy bueno/ a very good book in Spanish.")
In (145), because of the earlier mention of the speaker's diplomas, there is no need
to relate the antecedent "un diploma/a diploma" with another referent in the main clause.
The Object relative "[que me concedieron...]/[that they gave me...]" contains the Evoked
referent of'me/me,1 which refers to the speaker. However, there is no need for Anchoring
since Proposition-Linking grounding was already achieved before the utterance of the RC.
(144) and (145) show that Proposition-Linking grounding is favored by the use of
antecedents with the role of Predicate Nominal because these equational constructions
typically contain a subject and a predicate nominal that refer to the same entity in the
discourse.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
107
Similarly, the TC 'ex-s/ which also tends to strongly favor Proposition-Linking
grounding, has a type o f structure that typically does not favor any Grounding done by
Human Evoked referents, as can be observed in (146):
(146) EX-S
Enc.-i,...hay unas que me imagino que llegan a inscribirse o...y ya nunca vuelven?
Inf.-Hay muchas (enfermeras voluntarias)[que ni siquiera acaban el curso.] (Lope
Blanch, El Habla de la Ciudad 92)
'Enc.-are there some that I imagine that register or...they do not come back
anymore?
Inf.-There are many (volunteer nurses)[Xha\ do not even finish the course.]
The indefinite antecedent in (146) is relevant in the conversation by means of its
link to what the interviewer asked in the immediately preceding context about volunteer
nurses following up with their instruction at the hospital, prior to their service. The
speaker's answer in (146) refers to many candidates who work at the hospital that do not
finish the preparatory course. The Existential antecedents (Human and Nonhuman) are in
most of the cases Indefinite, as in (146). In the TC 'ex-s,' 91% of the Human antecedents
are Indefinite and 83% of the Nonhuman antecedents are Indefinite.
It is predictable that the TC 'ex-s' favors Proposition-Linking grounding, since
neither the Existential structure nor the Subject relative typically contain other Human
subject referents (other than the antecedent) to perform grounding (Main-Clause
grounding or Anchoring).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 0 8
As I explained earlier in this chapter, when analyzing the TC 'ex-s' with Human
referents, Fox and Thompson (309) noticed that Existential antecedents tended to favor
Subject relatives rather than Object relatives. Subject relatives were preferred because
when Human referents need grounding, they "do not need to be related to (other) humans
to be grounded." (309) Human referents “tend to be grounded by being related to their
own activities, that is, to earlier predicates, as with a locative, or by proposition-
linking....” (309) These tendencies in the discourse cause the preference for Subject
relatives over Object relatives, “since no other NP in the clause is needed to accomplish
the grounding.” (Fox and Thompson 309) In my investigation, I have shown that both
Human and Nonhuman referents with the TC 'ex-s' tend to favor Proposition-Linking
grounding. This suggests that it is the combination o f grammatical roles rather than the
Humanness of an antecedent of the RC what has an influence on the way the speakers
choose to make referents relevant in the discourse. Neither the Existential syntactic role
of the antecedent nor the Subject relative typically contains referents that are different
from the antecedent that needs Grounding. For this reason, the TC 'ex-s' tends to be
associated with Proposition-Linking grounding rather than with Anchoring or Main-
Clause grounding, which involve the relationship between an Evoked referent and a
referent that needs Grounding.
The TC 's-s' also represents a situation in which the Grounding tends to be
performed by Proposition-Linking grounding rather than by an Evoked Subject pronoun
(Anchoring or Main-Clause grounding). (147) shows a Human antecedent and (148)
shows a Nonhuman antecedent:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
109
(147) S-S
Enc.- Bueno, y Ovan a dejar ...a los militares seguir estudlando aparte ya con
su bachiUerato?
Inf. A .-... aquel2 2 [ que ha pisado una escuela], sabe perfectamente bien que lo
mejor para el hombre es la luz de la cultura.... (Lope Blanch, El Habla de la
Ciudad 210)
'Enc.-Well, and are the soldiers going to be allowed to continue with their
studies beyond their bachelors?
Inf A.-... the soldier [that has gone through school], (he ) knows perfectly that the
best thing for men is the enlightenment from culture;...1
(148) S-S
Guanajuato fiie en un tiempo la segunda ciudad mas importante... por la cantidad
de plata y oro que habia... Pero tu sabes que la mina de plata [que se llamaba La
Valenciana] fue en un tiempo la mina....(Lope Blanch. El Habla de la Ciudad 49)
'Guanajuato was in the past the second most important city..., due to the amount
of silver and gold that existed there. But you know that the silver mine [that
was called La Valenciana] was once the mine....'
In (147) and (148), the antecedents are relevant in the discourse because of their
link to the preceding context (by Proposition-Linking grounding). In (147), the previous
mention o f the soldiers pursuing studies is what determines the relevance o f the antecedent
“The accent on 'aquel' did not figure in the transcript that I used.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 1 0
"aquel/the soldier" and its RC. In (148), the previous mention o f the amount of silver and
gold that existed in Guanajuato is what determines that the antecedent and the RC are
relevant in the conversation.
Subject relatives are more appropriate than Object relatives when there is
Proposition-Linking because with this type o f grounding there is no need for Human
subjects (different from the element that needs grounding) to provide grounding.
In conclusion, the data o f my investigation reveal that the grammatical role of
Direct Object (in the antecedent and/or within the RC) favors Anchoring and Main-Clause
grounding, which are the types of grounding that consist of Evoked Human Subjects
providing the grounding. With Anchoring and Main-Clause grounding, the elements that
provide the grounding are either Human or Nonhuman ( they are only in rare occasions
animals in my data). The data reveal that when there is Anchoring o f Nonhuman
antecedents, in 95% o f the cases the element within the RC that anchors the antecedent is
Human. Similarly, when there is Main-Clause grounding, in 90% o f the cases the element
in the main clause that grounds the antecedent is Human. This correlation is highly
significant (p<.029) and can be observed in Appendix 4. The same correlation revealed
that Human antecedents with Anchoring and Main-Clause grounding tend to also be
grounded by Humans. However, this correlation with Human antecedents was not
significant, also evident in Appendix 4.
The relationships between the referents that provide the grounding and the entities
being grounded (by Anchoring or by Main-Clause grounding) are typically ownership,
usage, or manipulation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
I l l
The grammatical roles of Predicate Nominal, Subject, and Existential tend to favor
Proposition-Linking grounding because they are structures that usually do not include
elements (apart from the referent that needs grounding) that can provide grounding.
Therefore, they are more frequently associated with Proposition-Linking; their grounding
occurs by their link to earlier propositions.
The data show that the general tendencies observed are the same for Human and
Nonhuman antecedents, which suggests that the TC has an influence on the speakers'
selection of methods to give relevance to referents in the discourse.
3.3.3. Functional Role of the RC
My third hypothesis is that the different TCs (with Human and Nonhuman
antecedents) have an influence on the types of Functional Roles that the RCs have, and
that these preferences are determined by the type of Grounding.
Fox and Thompson found that in English the function of the RC o f Identifying a
referent is generally associated with Anchoring (303). This tendency was observed with
Nonhuman antecedents with the role of Subject, and Object relatives. Subject relatives in
English (Fox and Thompson 305-07) tend to favor Characterizations rather than
Identifications with Nonhuman antecedents that have the role of Direct Object.
Nonhuman antecedents with the role of Direct Object tend to favor Main-Clause
grounding. For this reason, Subject relatives tend to Characterize their already grounded
antecedents rather than Anchoring them, because “there is no communicative need to
ground a referent more than once.” (Fox and Thompson 305)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 1 2
The tendency observed by Fox and Thomson for Subject relatives to perform
Characterizations rather than Identifications was also observed in my data. In this section,
I will demonstrate that Subject relatives in Spanish tend to favor the function of
Characterization when there is Proposition-Linking grounding. This preference may be
explained by the property of Proposition-Linking of not being achieved by means of
Human referents grounding other (Human and Nonhuman) referents.
As was observed in the previous section, Proposition-Linking grounding tends to
be favored by RCs in which Human and Nonhuman referents are related to themselves,
and Characterizations are precisely the kind of mechanism which refers to referents as
Subjects of their own clauses, in terms of their own characteristics and habitual features
and activities, as illustrated in (149):
(149) ...,y en ese Iugar era donde Ios judios, en la noche, bajaban a tom ar agua de un
rio. [que pasaba por Jerusalem....] (Lope Blanch, El Habla de la Ciudad 190)
...,and in that place was where the Jews, at night, used to come down to drink
w ater from a river, [that passed by Jerusalem....]'
The Subject relative in (149) provides a descriptive feature of the antecedent: the
river in question passed by Jerusalem.
With Anchoring, Subject, Object, and Circumstantial Complement relatives tend to
favor the function of Giving New Information about an antecedent. The RCs that perform
Anchoring tend to be a device to relate Evoked referents (the anchors within the RC) to
Brand-new information in the RC in order to achieve the function of the RC of Giving
Brand-new information about an antecedent, as illustrated in (150):
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
113
(150) Una casa [que (yojvi ayer] fue puesta en venta hoy
'A house [that I saw yesterday] was put on sale today1
The RC in (150) anchors its antecedent by relating the Evoked referent o f the
pronoun "yo/I" to the antecedent "una casa/a house." Besides making the antecedent
relevant (by Anchoring), the RC also has the functional role of Giving New Information
about such antecedent. In (150), the entire RC provides Brand-new information about the
antecedent "una casa/a house."
When there is Main-Clause grounding, the tendency is also for Subject, Object,
and Circumstantial Complement relatives to Give New Information about already
grounded antecedents.
Proposition-Linking grounding, as I will show later, tends to favor
Characterizations more than the other two types o f grounding, and these Characterizations
are more commonly associated with Subject relatives. Proposition-Linking is a type of
Grounding in which referents that need Grounding do not have to be related to Human
(Evoked) referents (as with Anchoring and Main-Clause grounding). Therefore, Subject
relatives are more appropriate than Object or Circumstantial Complement relatives,
because in a typical Subject relative the coreferent o f the antecedent o f an RC is the
Subject of its RC. Being a subject in its own RC, a referent can be described in terms of
its own characteristics or habitual activities, rather than being referred to in terms of other
referents (subjects) in the discourse.
In order to show the distribution of Subject, Object, and Circumstantial
Complement relatives in the different functional roles of RCs, I establish correlations of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
114
the 11 TCs by functional role o f the RC for Human and Nonhuman antecedents, as shown
in Tables 5 and 6, respectively:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
TABLE 5
FUNCTIONAL ROLE OF THE RELATIVE CLAUSE BY TC FOR HUMAN ANTECEDENTS
FUNCTIONAL ROLE OF THE
RELATIVE CLAUSE
TYPE OF COMBINATION
OP-S PN-S DO-S DO-DO
N % N % N % N %
Characterization o f an antecedent 62 43% 43 28% 68 48% 1 6%
Identification o f an antecedent 15 10% 30 20% 9 6% 3 18%
Give New Information obout antecedent 54 38% 65 42% 56 40% 9 53%
Reaffirm the Existence o f an Antecedent 13 9% 16 10% 8 6% 4 24%
Total 144 100% 154 100% 141 100% 17 100%
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
TABLE 5 (Continued)
FUNCTIONAL ROLE OF THE
RELATIVE CLAUSE
TYPE OF COMBINATION
S-S S-DO EX-S TOTAL
N % N % N % N %
Characterization o f an antecedent 65 30% 1 6% 43 43% 283 36%
Identification o f an antecedent 39 18% 7 41% 2 2% 105 13%
Give New Information obout antecedent 85 39% 6 35% 44 44% 319 40%
Reaffirm the Existence o f an Antecedent 29 13% 3 18% 10 10% 83 11%
Total 218 100% 17 100% 99 100% 790 100%
p<,000, 4 cells (14.3%) have expected count less than 5,
116
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
TABLE 6
FUNCTIONAL ROLE OF THE RC BY TC FOR NONHUMAN ANTECEDENTS
FUNCTIONAL TYPE OF COMBINATION
ROLE OF THE
RELATIVE
CC-S c c - c c OP-S OP-DO PN-S PN-DO
CLAUSE
N % N % N % N % N % N %
Characterization
o f an Antecedent
51 48% 42 14% 33 23% 4 3% 57 31% 6 4%
Identification of
an antecedent
5 5% 49 16% 15 11% 26 17% 23 13% 47 30%
Giving New
Information about
an Antecedent
31 29% 160 53% 69 48% 69 45% 71 39% 71 45%
Reaffirming the
Existence o f an
Antecedent
19 18% 50 17% 26 18% 55 36% 32 18% 34 22%
Total 106 100% 301 100% 143 100% 154 100% 183 100% 158 100%
117
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
TABLE 6 (Continued)
FUNCTIONAL ROLE OF THE RELATIVE CLAUSE BY TC FOR NONHUMAN ANTECEDENTS
FUNCTIONAL TYPE OF COMBINATION
ROLE OF THE
RELATIVE
DO-S DO-DO S-S S-DO EX-S TOTAL
CLAUSE
N % N % N % N % N % N %
Characterization
o f an Antecedent
69 35% 8 4% 35 22% 4 3% 41 53% 350 19%
Identification of
an antecedent
17 9% 34 17% 22 14% 22 17% 4 5% 264 15%
Giving New
Information about
an Antecedent
77 39% 84 41% 81 50% 68 51% 28 36% 809 45%
Reaffirming the
Existence o f an
Antecedent
34 17% 77 38% 23 14% 39 29% 4 5% 393 22%
Total 197 100% 203 100% 161 100% 133 100% 77 100% 1816 100%
p<000
118
119
In Tables 5 and 6 it can be observed that Giving New Information is the functional
role most frequently achieved by the RCs overall. In Table 5 (Human antecedents) the
second functional role most frequently achieved by the RCs is Characterization (36%), but
in Table 6 (Nonhuman antecedents) Reaffirming the Existence o f an antecedent is the
second most preferred function (22%).
As I will explain later, the high frequency of Identifications with the TC 's-do'
(Human antecedents) is due to the high percentage of Evoked antecedents with this TC.
The TCs 'op-do,' 'do-do,1 and 's-do' with Nonhuman antecedents have high frequencies of
Reaffirming the Existence of an antecedent. This is probably due to the high frequencies
o f Inferable antecedents that these TCs have with Nonhuman antecedents.
In Table 5 ,1 only considered the TCs that did not produce empty cells. The TCs
that favor Giving New Information have either Subject or Object relatives, for instance:
the TC 'do-do' (53% of Giving New Information) and the TC 'pn-s' (42% of Giving New
Information). However, Characterizations are favored by TCs with Subject relatives, such
as 'op-s' (43%) and 'do-s' (48%). Notice that all of the TCs with Subject relatives have
high frequencies of Characterizations even if this is not their most favored type of
functional role.
The TCs in Table 5 with Object relatives present a different distribution, since they
are rare with Characterizations. The TCs 'do-do' and 's-do' each exhibit only 6% of
Characterizations. The high frequency of Identifications with the TC 's-do' is due to the
high percentage of Human antecedents with this TC that are Evoked (59%). As I
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 2 0
explained in Chapter 2, Identifications involve Evoked antecedents and Evoked
informational value within the RC.
The tendency observed in Table 5 (Humans) of Subject relatives having high
percentages of Characterizations can also be observed in Table 6 (Nonhumans). The main
pattern is for TCs with Subject relatives to favor Characterizations ('cc-s' (48%) and 'ex-s'
(53%)) or to favor Giving New Information, but show Characterization as a second
preference. This is the case of'do-s,' 'pn-s,' 'op-s,1 and 's-s,' which all have
Characterizations as a second preference, after the function of Giving New Information, as
Table 6 shows.
Similar to what was observed in Table 5, in Table 6 the TCs with Object relatives
show the lowest percentages o f occurrences with Characterizations. The TC 'cc-cc' also
shows a low frequency o f Characterizations. In Table 6 the preferences are for Giving
New Information, but the functional role o f Reaffirming the Existence of an antecedent
reaches high percentages, competing with the function o f Giving New Information. This
is the case with 'op-do' (36%) and 'do-do' (38%). This is expected, since, as I indicated
earlier, these two TCs tend to be Inferable with Nonhuman antecedents. The function of
Reaffirming the Existence o f an antecedent presupposes Evoked or Inferable antecedents
that are linked to earlier contexts.
In Table 6, the TC 's-do' does not exhibit a high percentage o f Identifications, as
was observed with Human antecedents in Table 5. This is due to the low percentage of
Evoked Nonhuman antecedents with the TC 's-do' (20%). In Table 6 the highest
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 2 1
percentage of Identifications is shown by the TC 'pn-do' (30%), which has 50% of Evoked
antecedents.
In Tables 5 and 6, the functions of Identification and Reaffirming the Existence of
an antecedent are predetermined by the Information Status o f the antecedents of the RCs.
Characterizations of Human and Nonhuman referents in my data may occur with Evoked,
Brand-New, or Inferable antecedents, but Evoked antecedents tend to favor them the
least. With Characterizations, RCs also tend to convey Brand-new and Inferable
information. In fact, only 2% of the Characterizations with Human antecedents provided
Evoked information in my data. With Nonhuman antecedents, only 1% conveyed Evoked
information.
Subject relatives are mechanisms to provide information about referents without
establishing relationships o f ownership, usage or manipulation, which are typical with
Object relatives. This can be seen in (151):
(151) DO-S
...puse en ...uno de mis primeros consultorios, un rotulo como de seis metros de
largo, grandote. enorme. [que decia: "Doctor Pablo B.; oidos, nariz,
exclusivamente... consultas....] (Lope Blanch, El Habla de la Ciudad 101)
'I put ...in one o f my first medical offices, a sign of about six meters long, very big.
[that said: "Doctor Pablo B.; ears, nose....]'
In (151), the Subject relative "[que decia:.,.]/[that said]" describes the sign that the
speaker had placed in front o f his medical office. The Subject relative is used by the
speaker to refer to the antecedent's characteristics in terms of its own characteristics.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 2 2
In Tables 5 and 6 it was observed that both Subject and Object relatives have high
percentages o f Giving New Information about Human and Nonhuman antecedents. (152)
and (153) illustrate that both Subject and Object relatives are appropriate for the
functional role of Giving New Information (not Characterization) about an antecedent:
(152) DO-S
Entonces, yo ya ...empece a meter un albanil. un peon, [que me empezara a
desmontar...a quitar la piedra] (Lope Blanch, El Habla Popular 191)
'Then, I started bringing a bricklayer, a worker, [that would start removing the
stone for me]
(153) DO-DO
Y una vez fui yo para recoger algunas...este...unas cosas [que yu habia dejado
alia....] (Lope Blanch, El Habla Popular 72)
'And once I went to pick up some— some things [that I had left there....]'
In (152), the Subject relative gives the hearer more (Brand-new) information about
an antecedent; that is, it communicates the type of work that the worker did for the
speaker. In (153), the coreferent of the antecedent within the RC is the Direct Object of
the personal pronoun 'yo/I.' Even though the coreferent of'unas cosas/some things' is not
the Subject of its own RC, the hearer is able to obtain Brand-new information about it,
that the things had been previously left by the speaker himself. In (153), unlike what was
observed in (152), the Brand-new information provided by the RC is conveyed in terms of
a relationship of manipulation of an object by a subject.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
123
The second part of my third hypothesis is that patterns of distribution o f the 11
TCs correlate with grounding and with different types o f functions of the RCs. The
analysis o f the factor Grounding is also useful to explain the general patterns observed
above.
In Table 7 1 established a correlation of the TCs (Human and Nonhuman
antecedents together) when they occur with Anchoring with the different types of
functional roles that the RCs have with respect to their antecedents. In this correlation I
do not include ail the TCs because some of them caused empty cells in the correlation.
The TCs ’ pn-do,' 'do-do,' and 's-do' did not show any cases of Characterization. I also
excluded the TC 'cc-s,' 'pn-s,' 'ex-s,' and 'do-s.' The TCs 'pn-s,' 'cc-s,' and 'ex-s' did not
show any case of Identification.
The two functional roles that have the highest percentages in Table 7 are Giving
New Information and Identification.
Ail the TCs favor Giving New Information, but the TCs with Subject relatives
('op-s' and 's-s') are the ones in which the function of Giving New Information has the
highest percentage (70% for 's-s' and 65% 'op-s') at the expense of the other functional
roles. These tendencies can be appreciated in Table 7, in which I show the correlation of
Functional Role of the RC by the three types of RCs, Subject relatives (which includes the
TCs with Subject relatives). Object relatives (includes the TCs with Object relatives), and
Circumstantial Complement relatives (the TC 'cc-cc') when there is Anchoring of Human
and Nonhuman antecedents together.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
124
TABLE 7
FUNCTIONAL ROLE OF THE RELATIVE CLAUSE BY TYPE O F RELATIVE
CLAUSE BY ANCHORING (HUMAN AND NONHUMAN ANTECEDENTS)
Functional Role
o f the Relative
Clause
Type of Relative Clause Total
Circumst.
Complement
Relative
Object
Relative
Subject
Relative
N % N % N % N %
Characterization 3 3% 2 1% 18 14% 23 6%
Identification 32 28% 41 24% 11 9% 84 21%
Giving New
Information
65 57% 90 54% 89 71% 244 60%
Reaffirming
Existence
of Antecedent
14 12% 35 21% 8 6% 57 14%
Total 114 100% 168 100% 126100% 408 100%
p<.000.
In Table 7 it is observed that Anchoring does not determine large differences of
functional roles of RCs for the types of RCs that are being analyzed. The function of
Giving New Information about an antecedent is the one that is clearly favored by all of the
TCs.
A clear pattern in Table 7 is that Subject relatives do not tend to Identify their
antecedents. In fact, only 9% o f the Subject relatives have the Identification function.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
125
Subject relatives are the most preferred mechanism to provide New information about an
antecedent. (154) shows a case of Giving New Information with a Subject relative when
there is Anchoring:
(154) S-S
Y entonces otro muchacho mexicano [que estaba ahi con nosotros], dijo que
estaba bueno.... (Lope Blanch, El Habla de la Ciudad 188)
'And then another Mexican guv [that was there with us], said that it was okay....'
In (154), a Subject relative gives Brand-new information about its antecedent. The
information contained within the RC is Brand-new, as it is the antecedent of the RC. In
this case the element that anchors the antecedent is not a Subject pronoun. Observe that
before the utterance of the Evoked pronoun "nosotros/us," there is a locative (Evoked),
"ahi," which refers to the place where the speakers were: Spain. This locative is the
element inside the RC that anchors the antecedent.
The RC in (154) does not perform Identification. Although it is surprising that the
speaker used "otro/another" for a Brand-new referent, there was not an earlier reference
to another guy.
(155) illustrates a Direct Object relative that Anchors and identifies at the same
time its antecedent:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
126
(155) S-DO
Nada mas el unico problema [que tuvimos] fiie lo de las bicicletas.... (Lope Blanch,
El Habla Popular 320).
'The only problem [that we had] was the thing with the bicycles....1
In (155), both the antecedent and the information within the RC are Evoked, since
they were mentioned earlier in the conversation. Before the utterance of (155), the
speaker had mentioned that the only problem that they experienced in the trip he was
talking about was the thing with the bicycles. Therefore, the mention in (155) involves an
Evoked referent and Evoked informational value within the RC.
The element within the RC that anchors the antecedent is the Evoked Subject
pronoun "nosotros," which is implicit within the RC. The presence of an implicit Subject
pronoun within a RC, which is not coreferent with an antecedent, is typical of an Object
relative, as it is with Circumstantial Complement relatives.
Anchoring with Object relatives and Circumstantial Complement relatives that
Give New Information about an antecedent is illustrated in (156) and (157):
(156) OP-DO
Enc.-JComo es?
In f-Bueno, muy diferente de lo [que yo me imaginaba], porque.... (Lope Blanch,
El Espanol 145)
Enc.-How is it?
Inf.-Well, it is very different from the thing [that I imagined], because....
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
127
(157) CC-CC
En una ocasion ya...por el ano de cuarenta v...cuatro...ren que (yo)hice un viaje a
los Estados Unidos....] (Lope Blanch, El Habla de la Ciudad 101)
'In one occasion already around the year forty four... [in which I made a trip to the
United States....]1
In (156) and (157) it can be observed that Object and Circumstantial Complement
relatives are both appropriate mechanisms to provide Brand-new information about an
antecedent. In each, the speaker conveys Brand-new information for the hearer in the
conversation, but this Brand-new information is related to Evoked Subject pronouns that
refer to the speaker within the RC.
Lambrecht has noted that “sentences typically contain some lexical or grammatical
manifestation o f the information assumed to be already given in the hearer's mind, as a
verbal point o f departure or basis for the new information to be added.” (51) That
observation can be applied to the cases seen above, since Anchoring, which involves
Evoked information in the discourse within the RC, is so clearly associated to the
functional role o f Giving New information. For instance, in (156) and (157), Evoked
information and Brand-new information interact within the RC, and the full RCs provide
Brand-new information for the hearer.
The correlation with Humans only produced a high number o f empty cells, due to a
low number of cases of Anchoring o f Human antecedents (76 cases). Although the
correlation was significant (p<044), every TC showed at least one empty cell.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1 2 8
In order to avoid empty cells, I joined all TCs with Subject relatives under one
variant, and did the same for all TCs with Object relatives. However, the correlation
including these two variants and the TC 'cc-cc' also showed empty cells with Human
antecedents.
I made the correlation with Nonhuman antecedents, which although being
statistically significant (p<001), exhibited several empty cells. For this reason, I made the
same correlation with type of RC (Subject, Object, and Circumstantial Complement
relatives). This correlation showed that the most favored Functional Role of the RCs with
Anchoring o f Nonhuman antecedents is Giving New Information about an antecedent, as
seen in Table 8:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
129
TABLE 8
FUNCTIONAL ROLE OF THE RELATIVE CLAUSE BY TYPE OF RELATIVE
CLAUSE BY ANCHORING (NONHUMAN ANTECEDENTS)
Functional Role Type of Relative Clause
of the Relative
Clause
CC
Relative
Subject
Relative
Object
Relative
Total
N % N % N % N %
Characterization 3 3% 6 9% 2 1% 11 3%
Identification 31 27% 4 6% 33 22% 68 21%
Giving New
Information
about an
Antecedent
65 58% 53 77% 83 55% 201 61%
Reaffirm the
Existence o f an
Antecedent
14 12% 6 9% 32 21% 52 16%
Total 113 100% 69 100% 150 100% 332 100%
p< 000. 3 cells (25.0%) have expected count less than 5.
The pattern observed in Table 7 can also be observed in Table 8, which is the
tendency for Subject relatives, Object relatives, and the TC 'cc-cc' to favor the functional
role of Giving New Information. Subject relatives (77%) favor this function more clearly
than Object relatives (55%) and the TC 'cc-cc' ( 58%). An interesting similarity between
Tables 7 and 8 is regarding Identification. Subject relatives only perform Identification in
6% of the cases, compared to other types o f functional roles. This indicates that
Identifications tend to trigger RCs in which typically there is a Subject Pronoun different
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
130
from the antecedent, and this pronoun can Identify such antecedent. For this reason
Subject relatives, which typically do not have this configuration, are less suitable for the
task of identifying an antecedent.
I made a correlation of the TCs with the different functional roles o f the RCs when
there is Main-Clause grounding (Human and Nonhuman antecedents). In this correlation I
did not include the TCs 'pn-do,' 'pn-s,' 'cc-s,' and 'ex-s' because they created empty cells.
What the TCs 'pn-s,' 'cc-s,' and 'ex-s' have in common is that they did not show any case of
Identification o f the antecedent. This reveals a tendency (already seen with Anchoring)
among Subject relatives with Main-Clause grounding towards not favoring Identification.
The TCs with Subject relatives ('op-s,' 's-s,' and 'do-s') tend to have low percentages of
Identification in comparison to the frequencies that they have with other functions (7%,
15%, and 9%).
The TCs with higher percentages o f Identifications are 'op-do' (29%) and 'do-do'
(32%). In order to observe general patterns in terms of Subject, Object, and
Circumstantial Complement relatives, I did a correlation of Functional Role o f the RC by
type of RC when there is Main-Clause grounding (Human and Nonhuman antecedents
together). This correlation is shown in Table 9:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
131
TABLE 9
FUNCTIONAL RO LE OF THE RC BY TC W IT H MAIN-CLAUSE GROUNDING
(HUMAN AND NONHUMAN ANTECEDENTS)
Functional Role o f Type of Relative Clause
the Relative
Clause
Circumstantial
Complement
Relative
Subject
Relative
Object
Relative
Total
N % N % N % N %
Characterization
o f an Antecedent
21 19% 169 39% 9 6% 199 29%
Identification of
an antecedent
9 8% 33 8% 40 27% 82 12%
Giving New
Information about
an Antecedent
63 58% 182 42% 75 51% 320 46%
Reaffirming the
Existence o f an
Antecedent
16 15% 50 12% 22 15% 88 13%
Total 109 100% 434100% 146100% 689100%
p< 000.
The tendency for Object relatives to have high frequencies of Identifications is seen
in Table 9. Notice that Object relatives reach a 27% of Identification. Subject relatives
reach a very high percentage in the functional role o f Characterization (39%), which is
very close to the percentage of Subject relatives for Giving New Information about an
antecedent (42%). The percentage o f Subject relatives in the function of Characterization
is higher with Main-Clause grounding than with Anchoring (14% of the cases with Subject
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
132
relatives perform Characterizations in Table 7). As was observed with Anchoring earlier
in this chapter, the functional role of Giving New Information about an antecedent is also
the preferred function with Main-Clause grounding. (158) and (159) show cases of Main-
Clause grounding in which the RC gives New Information about its antecedent:
(158) OP-DO
...el tuvo una reaction contraria a lo [que yo pensaba], y me indico.... (Lope
Blanch, El Habla de la Ciudad 17)
H e had a reaction that was contrary to the thing [that I was expecting], and he
indicated to me....'
(159) DO-S
Entonces, yo ya ...empece a meter un albanil, un peon, [que me empezara a
desmontar...a quitar la piedra] (Lope Blanch, El Habla Popular 191)
'Then, I started bringing a bricklayer, a worker, [that would start removing the
stone for me]
In (158), the Evoked referent is "el/he," which refers to the speaker's professor,
who was mentioned earlier. This is the element that grounds the antecedent of the RC.
The information contained within the RC is also Brand-new in the discourse.
In (159), the referent o f the pronoun "yo/I" is the element that performs Main-
Clause grounding. The information contained in the RC is Brand-new in the conversation.
In (158) and (159), Brand-new information is introduced by the RCs, but this
Brand-new information is related to Evoked elements within the RC (Evoked pronouns) in
order for the entire RC to provide Brand-new information about the antecedent..
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
133
As I explained above, Subject relatives reach a very high frequency o f
Characterizations in Table 9. (160) shows a case o f Characterization of an antecedent that
has been grounded by Main-Clause grounding:
(160) DO-S
Entonces, (el) encontro a una muchacha [que era huerfana tambien de papa y
mama.] (Lope Blanch, El Habla Popular 22 )
'Then, he found a girl [that was also an orphan.]'
In (160), the antecedent is grounded in the main clause, but because it is Indefinite,
the RC contains a characterization of a feature of the referent of the antecedent: the girl
was also an orphan.
The increase of Characterizations by Subject relatives with Main-Clause grounding
(with respect to Anchoring) may occur because when a Main-Clause occurs, an
antecedent has already received grounding by the time the RC is produced. For this
reason, there is no need to relate the referent of the antecedent to other referents within
the RC, but to the referent's own characteristics and activities, without establishing a
relationship of ownership or manipulation. This was noticed by Fox and Thompson (305-
07) with Nonhuman antecedents with the role of Direct Object, as I will discuss later in
this chapter.
As I did with Anchoring, I established correlations separating Human and
Nonhuman antecedents for the 11 TCs when there is Main-Clause grounding. In the
correlation with Humans, there were not enough cases for several TCs. For this reason,
most o f the TCs exhibited empty cells. A later correlation excluding these TCs did not
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
134
result in significant findings (p< 570). As I did in Table 8 with Anchoring, I joined all the
TCs with Subject relatives, and did the same with all the TCs that have Object relatives.
Table 10 shows the correlation o f Subject relatives and Object relatives by Functional Role
of the RC when there is Main-Clause grounding of Human antecedents (there were no
cases of the TC 'cc-cc'):
TABLE 10
FUNCTIONAL ROLE OF THE RC BY TYPE OF RELATIVE CLAUSE WITH
MAIN-CLAUSE GROUNDING (HTJMAN ANTECEDENTS)
Functional Role of Type o f Relative Clause
the Relative
Clause
Subject
Relative
Object
Relative
Total
N % N % N %
Characterization
of an Antecedent
77 41% 1 7% 78 39%
Identification of
an antecedent
13 7% 6 43% 19 10%
Giving New
Information about
an Antecedent
87 47% 6 43% 93 47%
Reaffirming the
Existence of an
Antecedent
9 5% 1 7% 10 5%
Total 186 100% 14 100% 200 100%
P<000. 2 cells (25.0%) have expected count less than 5.
In Table 10 the overall preference (47%) is for Subject and Object relatives to
perform the function o f Giving New Information of Humans.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
135
Object relatives have the same frequency for Giving New information (43%) and
for Identification (43%). Subject relatives in Table 10 only exhibit 7% o f Identifications,
but are very frequent with Characterizations (41%).
Although the function of Identification is predetermined by the Evoked
Information Status o f an antecedent (and by the Evoked information conveyed by the RC),
it is interesting that Identifications tend to be made by Object relatives rather than by
Subject relatives.
The same tendencies for Humans with Main-Clause grounding is observed with
Nonhumans grounded by this same type of grounding; Subject relatives and Object
relatives tend to Give New Information, but Subject relatives exhibit the highest
frequencies of Characterizations, while Object relatives are associated with high
percentages of Identifications.
Before doing the correlation with Subject, Object, and Circumstantial
Complements, a correlation with all TCs (the ones that did not exhibit empty cells) was
completed. This correlation showed that the TCs with Subject relatives ('op-s,' 'do-s,' and
's-s') exhibit high percentages of Characterizations (with respect to other functions), in
comparison to the TCs with Object relatives ('op-do,' 'do-do,' and 's-do'), and with
Circumstantial Complement relatives. Object and Circumstantial Complement relatives
showed low frequencies o f Characterizations with respect to the percentages that each of
them reached for other functional roles.
The TCs in which the Functional Role of Identification reaches its highest
frequencies have Object relatives. This is the case o f the TCs 'op-do' (26%) and 'do-do'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
136
(31%). The function o f R eaffirm ing the Existence of an antecedent reached its highest
percentage with the TC 'op-do' (26%).
In Table I I I summarize the tendencies explained above:
TABLE 11
FUNCTIONAL ROLE OF THE RC BY TYPE OF RELATIVE CLAUSE WITH MAIN-CLAUSE
GROUNDING (NONHUMAN ANTECEDENTS)
Functional Role Type of Relative Clause
of the Relative
Clause
Circumstantial
Complement
Relative
Subject
Relative
Object
Relative
Total
N % N % N % N %
Characterization
of an Antecedent
21 19% 92 37% 8 6% 121 25%
Identification of
an antecedent
9 8% 20 8% 34 26% 63 13%
Giving New
Information
about an
Antecedent
63 58% 95 38% 69 52% 227 46%
Reaffirming the
Existence of an
Antecedent
16 15% 41 17% 21 16% 78 16%
Total 109 100% 248100% 132100% 489 100%
p<000.
Table 11 clearly shows that the preferred function is Giving New Information
about an antecedent, but that Subject relatives strongly favor Characterizations, while
Object relatives are the type of RC most frequently associated with Identifications.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
137
Fox and Thompson (305-07) noted that with Nonhuman Direct Object antecedents (in
which Indefinite antecedents predominated), the fact that there was Main-Clause
grounding determined that the RC tended to characterize (using Subject relatives) its
antecedent already grounded in the main clause. Fox and Thompson only studied
Nonhuman referents as Direct Object antecedents. It is interesting that the pattern
observed in English for Nonhuman antecedents was observed in my investigation with
Human antecedents, because this result suggests that the factor Humanness may be not
triggering Characterizations . The TC 'do-s' with Main-Clause grounding of Human
antecedents tended to have Indefinite antecedents (55% of the antecedents are Indefinite).
The TC 'do-s' also showed a high frequency of Characterizations (48%). However, this
was not observed with Nonhuman antecedents, which tended to have Definite antecedents
(60% of the antecedents were Definite). The TC 'do-s' with Nonhuman antecedents
exhibited a lower percentage of Characterizations (33%). This suggests that the Indefinite
feature may be the factor that triggers a higher frequency o f Characterizations by the RC
when the antecedent has been grounded in the main clause, at least with the TC 'do-s'.
In Table 1 2 ,1 show the correlation of all the TCs (Human and Nonhuman) with
the different types of functional roles of the RCs when there is Proposition-Linking
grounding:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
TABLE 12
FUNCTIONAL ROLE OF THE RELATIVE CLAUSE BY TC BY PROPOSITION-LINKING GROUNDING (HUMAN AND
NONHUMAN ANTECEDENTS)
FUNCTIONAL ROLE TYPE OF COMBINATION
OF THE RELATIVE
CLAUSE
CC-S CC-CC OP-S OP-DO PN-S PN-DO
N % N % N % N % N % N %
Characterization 19 40% 17 23% 49 38% 1 1% 97 31% 5 3%
Identification 3 6% 8 11% 19 15% 5 7% 53 17% 51 34%
Giving New Information
about an Antecedent
13 28% 30 40% 43 33% 29 40% 119 38% 58 39%
Reaffirming Existence o f
Antecedent
12 26% 20 27% 19 15% 37 51% 47 15% 35 24%
Total 47 100% 75 100% 130 100% 72 100% 316 100% 149 100%
00
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
TABLE 12 (Continued)
FUNCTIONAL ROLE OF THE RELATIVE CLAUSE BY TC BY PROPOSITION-LINKING GROUNDING (HUMAN AND
NONHUMAN ANTECEDENTS)
FUNCTIONAL ROLE TYPE OF GROUNDING TOTAL
OF THE RELATIVE
CLAUSE
DO-S DO-DO S-S S-DO EX-S
N % N % N % N % N % N %
Characterization 57 43% 3 3% 72 29% 3 6% 67 51% 390 27%
Identification 8 6% 5 4% 41 16% 8 17% 6 5% 207 14%
Giving New Information
about an Antecedent
47 35% 46 38% 92 37% 16 34% 49 37% 542 37%
Reaffirming Existence
o f Antecedent
21 16% 67 55% 45 18% 20 43% 10 8% 333 23%
Total 133 100% 121 100% 250 100% 47 100% 132 100% 1472 100%
p<.000
u>
V O
140
In Table 12, the TCs that favor Characterizations have in common the fact that
they tend to be TCs with Subject relatives. A pattern in Table 12 that was not observed
with Anchoring or Main-Clause grounding is that the function of Reaffirming the
Existence is preferred by the TCs 'op-do,' 'do-do,' and 's-do,' which have in common the
fact that they have Object relatives. Table 12 exhibits different patterns than the
correlations seen so far in this chapter, in terms of the association of the TCs with the
different types of functional roles o f the RCs. The separation between Human and
Nonhuman antecedents will help explain these distributions, as I will show in what
follows, but it is interesting to note that, unlike what has been observed with Anchoring
and Main-Clause grounding, in Table 12 the functional role of Giving New Information is
not the most favored by all the TCs. This function is only favored by the TCs 'cc-cc'
(40%), 'pn-do' (39%), 's-s' (37%), and 'pn-s' (38%).
The TCs that tend to favor Characterizations are 'op-s' (38%), 'do-s' (43%), 'cc-s'
(40%), and 'ex-s' (51%). The TCs 's-s' and 'pn-s,' although favoring the function of Giving
New Information, also have high percentages of Characterizations (29% and 31%
respectively) that are not much lower than those for Giving New Information. The TC
'pn-do,' although favoring the function of Giving New Information, also has a very high
percentage of Identifications (34%).
The use of Subject relatives is influenced by the tendency, with Proposition-
Linking grounding, to talk about referents in terms of their characteristic features and
habitual activities. With this type of grounding, the coreferent of an antecedent tends to
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
141
be a Subject within the RC, since there is no need for other Human Subjects to ground an
antecedent by Anchoring or to ground by Main-Clause grounding.
The main difference between Table 12 and the Tables presented earlier with other
types of grounding is that 3 o f the 4 TCs with Object relatives in Table 12 favor the
function of Reaffirming the Existence o f an antecedent. This tendency for Object relatives
to favor the function of Reaffirming the Existence of an antecedent is summarized in Table
13:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
142
TABLE 13
FUNCTIONAL ROLE OF THE RC BY TYPE OF RELATIVE CLAUSE W ITH
PROPOSITION-LINKING GROUNDING (HUMAN AND NONHUMAN
ANTECEDENTS)
Functional Type of Relative Clause
Role o f the
Relative
Clause
Circum
stantial
Complement
Relative
Subject
Relative
Object
Relative Total
N % N % N % N %
Characteri
zation o f an
Antecedent
17 23% 361 36% 12 3% 390 27%
Identification
o f an
antecedent
8 11% 130 13% 69 18% 207 14%
Giving New
Information
about an
Antecedent
30 40% 363 36% 149 38% 542 37%
Reaffirming
the Existence
of an
Antecedent
20 27% 154 15% 159 41% 333 23%
Total 75 100% 1008 100% 389100% 1472 100%
p<.000.
Table 13 shows the high frequency (41%) o f Object relatives with Proposition-
Linking grounding that Reaffirm the Existence o f their antecedents.
In a first correlation separating Humans and Nonhumans, the correlation with
Human antecedents (p< 000, with 23 cells with an expected count of less than 5), resulted
in several empty cells.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
143
The TCs 'cc-cc' (2 cases), 'pn-do' (14 cases), 'op-do' (4 cases), 'do-do' (9 cases),
'cc-s' ( 3 cases), and 's-do' (3 cases) were for this reason excluded from a second
correlation.
Table 14 shows the correlation established for the 5 TCs that did not cause empty
cells for Human antecedents with Proposition-Linking grounding:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
TABLE 14
FUNCTIONAL ROLE OF THE RELATIVE CLAUSE BY TC BY PROPOSITION-LINKING GROUNDING
(HUMAN ANTECEDENTS)
FUNCTIONAL ROLE OF
THE RELATIVE CLAUSE
TYPE OF COMBINATION
OP-S PN-S DO-S S-S EX-S TOTAL
N % N % N % N % N % N %
Characterization 32 49% 40 28% 28 52% 47 32% 35 47% 182 38%
Identification 9 14% 30 21% 3 6% 27 18% 2 3% 71 15%
Giving New Information
about an Antecedent
14 22% 56 40% 19 35% 49 33% 30 41% 168 35%
Reaffirming Existence of
Antecedent
10 15% 15 11% 4 7% 26 17% 7 10% 62 13%
Total 65 100% 141 100% 54 100% 149 100% 74 100% 483 100%
P<000.
145
It is revealing that overall, Table 14 does not favor the Function o f Giving New
Information (35%), but the functional role o f Characterization (38%). In fact, Table 14
shows that with Proposition-Linking grounding, 3 o f 5 TCs in the correlation favor
Characterizations. Only the TC 'pn-s,' which has 28% o f Characterizations, does not favor
this function. The TC 's-s' has almost the same percentage for Characterization (32%) and
for Giving New Information (33%).
In Table 14 we see that Characterizations tend to be performed by Subject
relatives. To observe the interaction of Proposition-Linking grounding with the function
of Characterization o f an antecedent by a Subject relative, consider the examples in (161),
(162) and (163):
(161) OP-S
Ha salido tanto ingeniero, tanto del Poli, como de la Universidad....
...Osea que ahorita, para el ingeniero [que esta preparado] no hay dificultad....
(Lope Blanch, El Habla de la Ciudad 170)
'So many engineers have graduated either from the Poli or from the University....
...So right now, for the engineer [that is prepared] there is no difficulty....
(162) DO-S
Creo que si puede dividirse la mujer en las dos partes y poder satisfacer las dos
necesidades. El empleo, en si, te exige demasiado, y el hogar....
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
146
Inf.-Pero a una muier [que trabaja] y [que tiene hijos y tiene casa,] jamas en la vida
la veras neurastenica. (Lope Blanch, El Habla de la Ciudad 269).
1 think that women can be divided in the two parts and be able to satisfy both
necessities.
Your job demands a lot from you, and home....
Inf.- But a woman [that works] and [that has children and a home], you will never
see her neurotic'
(163) PN-S
Digo, no...no hables del termino medio mexicano o del termino medio...eh...de
nuestra...nuestra clase media de abajo...pero tambien el hombre es un...es una
persona [que no pasa de oficinista.] (Lope Blanch, El Habla de la Ciudad 262)
7 say, no ...do not talk about the average Mexican or about our lower middle
class...but also the man is a ...is a person [that is not more than an office worker.]
In (161), the descriptive feature (prepared engineers) within the RC can be inferred
from the earlier context, which mentions that several engineers have graduated from an
institution (the Poli). In (161), the RC provided a Characterization of the Human
antecedent "el ingeniero/the/an engineer."
In (162), each o f the two RCs provide a descriptive feature of the antecedent "una
mujer/a woman." In (163), the antecedent is relevant due to its link to the immediately
preceding context. The RC gives a descriptive feature, describing the profession o f the
average lower-middle class Mexican.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
147
In the examples above, it can be observed that Humans do not need other Human
referents in order to be characterized. The results above support Fox and Thompson's
(309) observations about Human referents. Human referents do not need to obtain
grounding by means o f their relationship with other Human referents. Instead, the
tendency is for Humans to relate to themselves, in terms of their activities. This
observation was made by Fox and Thompson when they analyzed the TC 'ex-s,1 in which
all the antecedents were Human. Subject relatives tend to favor the tendency for Human
referents to be related to themselves rather than being related to other Humans that
ground them, and this is compatible with the function o f Characterization and with
Grounding by Proposition-Linking.
The correlation of Functional role of the RC by type of RC by Proposition-Linking
grounding of Human antecedents was significant (p<001), and showed empty cells for
Object relatives and for Circumstantial Complement relatives.
In Table 15 1 establish the correlation of the functional roles of the RCs with the
11 TCs when there is Proposition-Linking of Nonhuman antecedents:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
TABLE 15
FUNCTIONAL ROLE OF THE RELATIVE CLAUSE BY TC BY PROPOSITION-LINKING GROUNDING
(NONHUMAN ANTECEDENTS)
FUNCTIONAL ROLE OF
THE RELATIVE CLAUSE
TYPE OF COMBINATION
CC-S CC-CC OP-S OP-DO PN-S PN-DO
N % N % N % N % N % N %
Characterization 19 43% 17 23% 17 26% 1 2% 57 33% 5 4%
Identification 3 7% 8 11% 10 15% 5 7% 23 13% 44 33%
Giving New Information
about an Antecedent
12 27% 28 38% 29 45% 27 40% 63 36% 52 39%
Reaffirming Existence of
Antecedent
10 23% 20 27% 9 14% 35 52% 32 18% 34 25%
Total 44 100% 73 100% 65 100% 68 100% 175 100% 135 100%
4 *
00
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
TABLE IS (Continued)
FUNCTIONAL ROLE OF THE RELATIVE CLAUSE BY TC BY PROPOSITION-LINKING GROUNDING
(NONHUMAN ANTECEDENTS)
FUNCTIONAL ROLE OF TYPE OF COMBINATION
THE RELATIVE CLAUSE
DO-S DO-DO S-S S-DO EX-S TOTAL
N % N % N % N % N % N %
Characterization 29 37% 3 3% 25 25% 3 7% 32 55% 208 22%
Identification 5 6% 4 4% 14 14% 7 16% 4 7% 127 13%
Giving New Information
about an Antecedent
28 35% 41 37% 43 43% 15 34% 19 33% 357 37%
Reaffirming Existence of
Antecedent
17 22% 64 57% 19 19% 19 43% 3 5% 262 28%
Total 79 100% 112 100% 101 100% 44 100% 58 100% 954 100%
P<.000
150
The main difference between Table 15 and the Tables discussed earlier is that 3 of
the 4 TCs with Object relatives tend to favor the function of Reaffirm ing the Existence of
an antecedent. This function is more frequent with Nonhuman antecedents (28%) than
with Human antecedents (13%) in Table 14.
The function o f Giving New Information, as observed with all types o f Grounding,
is favored by TCs with Subject and Object relatives, along with Circumstantial
Complement relatives (TC 'cc-cc').
Three o f the four TCs with Object relatives in Table 15 favor the function of
Reaffirming the Existence of an antecedent. In Table 15 the TCs ’ op-do,' 'do-do,' and's-
do1 favor the function of Reaffirming the Existence of an antecedent. It can be observed
that 'op-do' and 'do-do* are the TCs that exhibit the highest percentages of occurrences
with this function (52% and 57% respectively). This suggests that Nonhuman
antecedents tend to be reaffirmed or reestablished in the discourse by Human referents,
and that the most appropriate mechanism is an Object relative rather than a Subject
relative. (164) and (165) illustrate this functional role by the RC when there is
Proposition-Linking grounding of Nonhuman antecedents:
(164) DO-DO
...y alii Io fueron a encontrar a mi hermano, muerto.
Enc.-^Con balazos, o herido, nada?
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
151
Inf.-Pues fij'se que no tenia nada; pero...pus yo creo /'agua le desparecio-cverda?-
alguna herida [que (el) haiga tenido]; porque si alii se paso toda la noche.... (Lope
Blanch, El Habla Popular 174).
'...and they found my brother there, dead.
.Ewe.-With gunshots, or wounded, nothing?
/w/I-Well he did not have anything; but...I think the water vanished-you know?-
some wound [he may have had]; because he spent the whole night there....'
(165) OP-DO
...[necesitamos] mas... gente joven, educada, que venga y se mete y use las
palabras que estos americanos les gusta oir; una persona bien preparada, bien...de
too esto [que eos quieren....] (Lope Blanch, El Espanol 164)
'...[we need] more ...young people, educated, who will come and use the words
that these Americans like listening to; a well prepared person...of all of these things
[that Americans want....]'
In (164), the relevance of the antecedent is guaranteed by the previous mention of
the speaker’ s brother, who had been found dead. The interviewer asked whether the
speaker's brother had been wounded by gunshots. The information contained within the
RC does not give any Brand-new information, but reaffirms its antecedent by providing
Inferable information for the hearer. Notice that there is no Identification in (164) because
neither the antecedent nor the informational value within the RC are Evoked. Both the
antecedent and the informational value within the RC are Inferable from the earlier
discourse.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
152
Similarly in (165), the antecedent is made relevant by its link to the immediately
preceding context in which the speaker refers to the way Americans in the community
want to be spoken to by the Mexican-American community. The information contained
within the RC, '[que ellos quieren]/[that Americans want]1 is Inferable from the
immediately preceding context in which the speaker talked about the way Americans want
to discuss certain issues.
Object relatives are better suited than Subject relatives for the task of Reaffirming
the Existence of an antecedent because the referents of these antecedents are Nonhuman.
It has been observed earlier in this chapter that a typical Object relative has a syntactic
configuration that involves an Evoked Human Subject, which establishes a Subject-Object
relationship with a Human or a Nonhuman referent. Usually within an Object relative, the
relationship between a Human Subject an its Nonhuman Direct Object is one of
manipulation, ownership, or usage, as noted by Fox and Thompson (304). In both (164)
and (165), the informational value within the RC is Inferable because o f the earlier
statements made by the speakers in each example. The fact that Nonhumans rather than
Humans tend to be reaffirmed in the discourse suggests that Nonhuman referents tend to
be validated in the discourse by their relationships with Human referents. Within an
Object relative, it is very common to find a Subject pronoun that is Human. This Human
subjects are the element that tend to reaffirm Nonhuman referents, as seen in (164) and
(165).
But why does Proposition-Linking grounding favor this type of function so
clearly?. The answer is that the antecedents of the RCs that reaffirm the existence o f an
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
153
antecedent tend to be either Inferable or Evoked, and the informational value of the RC
when this function is performed is always Inferable. Proposition-Linking grounding means
that referents are made relevant by their relation with the earlier context. For this reason,
it is not surprising that it is this type o f grounding the one that tends to be more commonly
associated with the function of Reaffirming the Existence o f an antecedent that is be
Evoked or Inferable from the earlier discourse.
Finally, it is important to notice that in Table 15 there is a tendency for the TC 'pn-
do' to perform Identification o f its antecedent, as in example (166):
(166) Inf. A .-...Y ya. Nos pasaron para aca, y ya nos toco puest's.
Inf-A -Entons, este puesto, que es este [que me dieron]....(Lope Blanch, El Habla
Popular 367-8).
'Inf.A.~ And that was it. We were transferred here, and we got stands.
Inf.A.- Then, this stand, which is this one [that they gave me]...'
In (166), Proposition-Linking grounding occurs because the antecedent and its RC
are relevant due to their link to the preceding context, in which the speaker mentioned that
he and his friends obtained stands to sell their merchandise. In (166), the antecedent is
Evoked due to its previous mention ("este puesto/this stand.") The information contained
within the RC is also Evoked, since the speaker had previously mentioned that he and his
friends had gotten stands to sell their merchandise. An Object relative, in which typically a
Nonhuman referent is the Direct Object of a Human referent within the RC, is more
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
154
suitable than a Subject relative for the task o f identifying the Nonhuman referent in (166).
The Predicate Nominal role o f the antecedent is what allows the speaker to summarize the
actions described earlier for the hearer and introduce them after the copulative verb
"es/is." The Predicate Nominal role position o f the antecedent allows the Evoked referent
in (166) to be reintroduced in the conversation by the speaker. Fox and Thompson (302-
4) noticed the preponderance o f Object relatives for performing Anchoring and
Identification at the same time. This observation was made in their study of Nonhuman
antecedents with the TC 's-do.' Although I did not show the TC 's-do' with Humans that
receive Anchoring because it caused an empty cell, it is important to notice that
Identification was the preferred function o f this TC (46% of Identifications). Nonhumans
with TC 's-do' when there is Anchoring, on the other hand, preferred the function of giving
Brand-new information about their antecedents. The reason for this difference between
Humans and Nonhumans in my data is that Human antecedents with the TC 's-do' tend to
be Evoked (which favored Identification), whereas Nonhuman referents with the TC 's-do'
when there is Anchoring tend to be Inferable.
To summarize, these results show that the different TCs in this study are
associated with a number of functional roles performed by the RCs. The association of
Proposition-Linking grounding with various functional roles that these TCs have when
they have Human or Nonhuman antecedents was also demonstrated.
With Proposition-Linking grounding, Subject relatives are a suitable mechanism to
perform Characterizations of Humans and Nonhumans. Subject relatives favor
Characterizations because the structure o f a Subject relative is more appropriate than the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
155
structure o f an Object relative for the task o f referring to an entity in terms o f its own
features or habitual activities.
With the function of Reaffirming the Existence of an antecedent, antecedents are
either Inferable or Evoked, and the information conveyed by the RC is always Inferable.
For this reason, the association o f this function with Proposition-Linking is not surprising,
since this type of grounding usually involves Inferable and Evoked referents, since the
referents obtain relevance by means of their relation to the earlier discourse. The fact that
Object relatives are preferred for this function suggests a tendency for Nonhuman
referents to be validated or reinforced in the discourse by verbs whose subjects tend to be
Human and Evoked. This tends to trigger Object relatives rather than Subject relatives,
since the subjects of the verbs that reaffirm the existence o f an antecedent tend to be
Human and different from the antecedents.
Similarly, Identification obtained a high percentage with Proposition-Linking, as
was observed in Table 15 for the TC 'pn-do.' This result was not unexpected, since
Evoked antecedents and Evoked informational value within the RCs are requirements for
Identifications, and Proposition-Linking grounding tends to favor Inferable or Evoked
antecedents. The preference for Object relatives over Subject relatives suggests that
although the relevance o f an antecedent and its RC is obtained by their link to earlier
propositions in the discourse, the task o f identifying a referent involves a Human subject
pronoun that is not coreferent with the antecedent that is being identified.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
156
3.4. Conclusions
Humanness has an influence on the distribution o f syntactic roles both in the RC
and in its antecedent. While Human referents clearly tend to be coded as Subjects when
they are antecedents o f RCs, and as Subject relatives, Nonhuman referents are more
frequently coded as Direct Object and Circumstantial Complement antecedents, but they
can be relativized as Subjects or Direct Objects.
The predominance of the role of Subject (of the antecedent and o f the RCs) for
Humans and o f the role of Direct Object for Nonhuman antecedents, and the non
predominance o f Subject relatives over Object relatives for Nonhumans reveals that
relativization follows the general tendency observed in Spanish for Humans to be coded as
Subjects and for Nonhumans to be coded as Direct Objects (Silva-Corvalan, "On the
Interaction.")
In this chapter I have also shown the association of grammatical roles with types of
grounding. Object relatives are more commonly associated with Anchoring, because an
Object relative typically involves Evoked Human Subject pronouns that relate to
antecedents that need grounding. With Nonhuman antecedents, usually there is a
relationship o f ownership, usage, or manipulation between a Human Subject o f an Object
relative and the antecedent that is being grounded.
I have also shown that referents with Main-Clause grounding tend to be Direct
Objects rather than Subjects. This is due to the word order SVO that is typical in Spanish.
In the main clause, a Direct Object antecedent is typically grounded by a Human Evoked
Subject.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
157
The TCs that favor Proposition-Linking grounding do not tend to have
antecedents with the role o f Direct Object, as with Main-Clause grounding. These are the
TCs 'pn-s,' 's-s,' 'ex-s,' and 'pn-do.' The TCs that tend to favor Proposition-Linking
grounding tend to have Subject relatives. When there is Proposition-Linking grounding,
antecedents obtain relevance by their link with earlier contexts. Subject relatives are a
grammatical device that allows the coreferent o f the antecedent o f a RC to be the Subject
within its own RC. When grounding is accomplished by Proposition-Linking, the
antecedent does not need other referents to obtain grounding.
The TCs 'pn-do' and 'pn-s' usually have Human and Nonhuman Evoked and
Inferable antecedents. This tendency also explains the preference for Proposition-Linking.
Equational constructions such as predicate nominals do not tend to involve Brand-new
referents. Similarly, Proposition-Linking grounding does not tend to involve Brand-new
referents as antecedents of RCs.2 3
“In cases in which there was a Brand-New element in the antecedent, it
was coded as Brand-New even if there had been Proposition-Linking grounding,
as in the following example:
Hay muchos problemas de delicuencia en esta ciudad. Ayer hubo tres
casas en este vecindario [que fiieron victimas de robo].
'There are lots of problems with delinquency in this city. Yesterday there
were three houses in this neighborhood [that were robbery victims]'
In the example above, the antecedent would be coded as Brand-New since
the fact that there were three houses in this neighborhood is Brand-New for the
hearer. The information within the RC is Inferable from the earlier mention of
delinquency problems. In this case, the RC grounds its antecedent by its relation
with the earlier context.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
158
The correlation o f Functional Role o f the RC by type of RC for Human and
Nonhuman antecedents can be summarized in Tables 16 and 17:
TABLE 16
FUNCTIONAL ROLE OF THE RELATIVE CLAUSE BY TYPE OF
RELATIVE CLAUSE (HUMAN ANTECEDENTS)
Functional Role of Type of Relative Clause
the Relative
Clause
Subject
Relative
Object
Relative
Total
N % N % N %
Characterization
of an Antecedent
285 37% 2 3% 287 35%
Identification of
an antecedent
95 12% 23 36% 118 14%
Giving New
Information about
an Antecedent
308 40% 28 44% 336 41%
Reaffirming the
Existence of an
Antecedent
78 10% 11 17% 89 11%
Total 766 100% 64 100% 830 100%
P<000.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
159
TABLE 17
FUNCTIONAL ROLE OF THE RELATIVE CLAUSE BY TYPE OF RELATIVE
CLAUSE (NONHUMAN ANTECEDENTS)
Functional Type o f Relative Clause
Role o f the
Relative
Clause
Circum
stantial
Complement
Relative
Subject
Relative
Object
Relative Total
N % N % N % N %
Characteri
zation o f an
Antecedent
42 14% 286 33% 22 3% 350 19%
Identification of
an antecedent
49 16% 86 10% 129 20% 264 15%
Giving New
Information
about an
Antecedent
160 53% 357 41% 292 45% 809 45%
Reaffirming the
Existence of
an Antecedent
50 17% 138 16% 205 32% 393 22%
Total 301 100% 867 100% 648100% 1816 100%
p<000.
Table 16 does not include Circumstantial Complement relatives because there were
only 3 cases of this type of RC with Human antecedents. 2 out of the 3 cases (67%) with
this type o f RC occurred in the function o f Giving New Information about an antecedent.
Tables 16 and 17 show that the function o f Giving New Information is the most favored
by Subject, Object, and Circumstantial Complement relatives. With Human antecedents,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
160
Subject relatives are also very frequent with Characterizations. This can be observed for
Humans and Nonhumans, but it is more clear among Human antecedents.
Object relatives have a high percentage of Identifications with Human antecedents
(36%), but with Nonhuman antecedents the percentage o f Reaffirming the Existence of an
antecedent is very high (32%). The results regarding Identification and Reaffirming the
Existence of an antecedent may also be due to differences o f Information Status of the
antecedents. Identifications always involve Evoked antecedents, whereas Reaffirming the
Existence o f antecedent presupposes Evoked or Inferable antecedents.
The preference for Object relatives over Subject relatives for the functions of
Identification and Reaffirming the Existence of an antecedent suggests that referents need
other referents that are different than them in order to be identified or reaffirmed in the
discourse. An Object relative typically contains a Subject pronoun (often Human) that has
the coreferent of the antecedent as the Direct Object of the RC. This Subject-Direct
Object relationship within the RC is compatible with the function of Identification and the
function of Reaffirming the Existence o f an Antecedent.
The correlation of the factor TC with Functional role of the RC when there is
Anchoring showed that Anchoring is a device that allows the speakers to provide Brand-
new information about a referent by relating Brand-new Information with an Evoked
referent (the anchoring element within the RC) in order for the RC to convey Brand-new
information. This pattern was observed with Human and Nonhuman antecedents.
With Main-Clause grounding (Human and Nonhuman antecedents), the function of
Giving Brand-new information about an antecedent was also the most favored function.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
161
Brand-new information is conveyed by RCs that do not need to ground their antecedents,
since grounding was already achieved in the main clause.
Proposition-Linking grounding among Humans tends to produce Subject relative
clauses that favored Characterization o f antecedents. This pattern, which is different from
what was observed with Anchoring and Main-Clause grounding, indicates that Subject
relatives in cases of Proposition-Linking grounding are the most suitable mechanism to
refer to Human antecedents in terms o f their own habitual activities and features, without
establishing relationships o f ownership or manipulation. Because referents do not need to
be related to other (Human) referents to obtain Proposition-Linking grounding, they can
be Subjects within their own RCs. This tendency was also observed with Nonhuman
antecedents. The preference for Subject relatives over Object relatives to refer to Human
referents that need grounding in terms of their own characteristics and activities was
previously observed by Fox and Thompson (1990: 309).
Although the most generally preferred function with Proposition-Linking
grounding of Nonhuman antecedents is not Characterization, my study has shown that
Nonhuman antecedents with Proposition-Linking grounding also tend to exhibit Subject
relatives with high frequencies of Characterizations.
Reaffirming the Existence of an antecedent involves a referent's link to an earlier
context, which is also a feature of Proposition-Linking grounding. For this reason, it is
not surprising that this functional role increases significantly with Proposition-Linking
grounding of Nonhuman antecedents.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
162
This increase of the function o f Reaffirming the Existence o f an antecedent can be
summarized in Table 18:
TABLE 18
FUNCTIONAL ROLE OF THE RELATIVE CLAUSE BY TYPE OF RELATIVE CLAUSE BY
PROPOSITION-LINKING GROUNDING (NONHUMAN ANTECEDENTS)
Functional Role of Type of Relative Clause
the Relative
Clause
Circum
stantial
Complement
Relative
Subject
Relative
Object
Relative Total
N % N % N % N %
Characterization
o f an Antecedent
17 23% 179 34% 12 3% 208 22%
Identification of
an antecedent
8 11% 59 11% 60 17% 127 13%
Giving New
Information about
an Antecedent
28 38% 194 37% 135 38% 357 37%
Reaffirming the
Existence of an
Antecedent
20 27% 90 17% 152 42% 262 28%
Total 73 100% 522 100% 359100% 954 100%
p<000.
Unlike Table 14 (Human antecedents with Proposition-Linking grounding), in
which Characterizations were the most preferred function, Table 18 shows that overall
with Nonhuman antecedents the second most preferred function (after Giving New
Information about an antecedent) is Reaffirming the Existence o f an antecedent (28%).
This function is favored by Object relatives, since 42% of the Object relatives with
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
163
Nonhuman antecedents have the function o f Reaffirming the Existence o f an antecedent
according to Table 18.
Object relatives are preferred over Subject relatives because they provide a device
by which a referent is validated or reaffirmed in the discourse by its relation with another
referent (usually a Subject pronoun).
The TC 'pn-do' tends to provide Identifications when there is Proposition-Linking
grounding o f Nonhuman referents, as Table 15 revealed. The Predicate Nominal position
of the antecedent, after the verb 'serT to be' makes it possible to establish the link between
the RC and its antecedent, and the information provided in the earlier discourse. The
information in the RC and in its antecedent are relevant because of the earlier discourse.
The results in this chapter suggest that relativization is influenced by factors that
are present in the interaction among the speakers. Keenan and Comrie's and Comrie and
Keenan's Accessibility Hierarchy for relative clause formation, unlike my study, is a
typological description o f the syntactic positions that could be relativized in a number of
languages, based on cognitive difficulty, which is to say some syntactic positions are easier
to relativize than others for a number of languages. On the other hand, my study has
focused on the semantic/pragmatic factors that correlate statistically with the production
of different syntactic roles when there is relativization. According to this,
semantic/pragmatic factors interact with the choices that the speakers make in terms of the
syntactic roles that are involved in the process of relativization.
My study has shown that Humanness is a factor that has an influence on the
speakers' selection of strategies o f relativization to refer to Human and Nonhuman entities
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
164
in their discourse. There is a very strong tendency for Humans to be relativized as
Subjects, but this tendency was not observed with Nonhuman referents, which tend to be
relativized either as Subjects or Objects. Because of the tendency for Humans to talk
about other Humans (Fox and Thompson 304), and given that Humans tend to be coded
as grammatical Subjects in Spanish (Silva-Corvalan, "On the Interaction"), it is expected
that the speakers develop a strategy when they relativize Human referents.
The nonpredominance o f Subject relatives over Object relatives when the
antecedents are Nonhuman may be due to the fact that Object relatives typically contain a
Subject pronoun in a Subject-Direct Object relationship within the RC. In this
relationship, the Nonhuman coreferent of the antecedent is an entity that is typically
manipulated, owned, or used. Fox and Thompson (304, citing Du Bois 269-70) have
noted that Nonhuman referents tend to obtain relevance in terms o f Human referents who
establish relationships of usage, ownership or manipulation with these Nonhuman
referents.
I have also shown in this chapter that the grammatical roles involved in
relativization correlate with the different ways in which speakers make referents relevant
in the discourse. Object relatives favor Anchoring of referents, whereas Subject relatives
tend to favor Proposition-Linking grounding.
Subject and Object relatives tend to favor the function of Giving New Information
about an antecedent, but Characterizations are favored by Subject relatives, whereas the
functions Identification and Reaffirming the Existence o f an antecedent tend to be favored
by Object relatives.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
165
In summary, I have demonstrated the importance of paying attention to the
interaction o f semantic, pragmatic, and functional factors when studying the grammatical
variation observed in relativization.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
166
CHAPTER 4
DISCOURSE AND SOCIAL VARIABLES
4.1. Introduction
In this chapter I analyze the interaction of social variables with the production of
Restrictive and Nonrestrictive RCs. Nonrestrictive RCs have, as I explained in Chapter 2,
an evaluative property, which consists of providing commentary about an already
specified antecedent. Consider, for instance, (167), which illustrates a Nonrestrictive
RC:
(167) ...en ese primer curso estudiabamos la Metafisica General, [que nos daba el
Pao...Pablo Deza] (Lope Blanch, El Habla de la Ciudad 59).
'...in that first course we studied General Metaphysics, [which was taught by
Pao...Pablo Deza]1
In (167), the name of the professor who taught General Metaphysics is conveyed
by the RC. The information conveyed by the RC is not required by the hearer in order to
locate the specific referent of the already specified antecedent. Instead, the
Nonrestrictive RC is conveying a comment about the referent of the antecedent. The
function o f the Nonrestrictive RC is not restricting, but it is instead evaluative with
respect to its antecedent. Following Labov's definition of an evaluation (Language 366)
as “the means used by the narrator to indicate the point of the narrative,” and applying it
to the analysis o f Nonrestrictive RCs, it can be said that the Nonrestrictive RC is used to
provide an evaluation or explanation o f its antecedent.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
167
In the present investigation I show that the more frequent use of Nonrestrictive
RCs by speakers with a higher level of education and higher socioeconomic level
suggests the existence o f different styles of communication among different social groups
in terms of the speakers' production o f RCs.
I also show the correlation of the Presence/Absence of a preposition with social
variables such as level of education and socioeconomic level. The results of my
investigation reveal that speakers with a higher level o f education and speakers from a
higher socioeconomic level tend to favor the Presence of a preposition with a relativizer.
Speakers with higher levels of education and speakers who belong to higher
socioeconomic levels are typically more familiar with the Spanish standard norm that
recommends the use o f a preposition before the relativizer "que/that." This may explain
the results mentioned above.
In the present study, monolingual speakers tend to produce the standard variant
(the Presence of a preposition) more frequently when they are compared to bilingual
speakers. As I will explain below, the bilingual speakers' lack of familiarity with the
Spanish normative grammar may explain their infrequent production of a preposition
with a relativizer in contexts in which the variable use of a preposition with a relativizer is
possible in Spanish. As explained in Chapter 2, the use of a preposition with the
relativizer "que/that" is mandatory according to Real Academia Espanola (529) when the
relativized position is Circumstantial Complement. In the present investigation, the
majority of cases in which the variable use of a preposition with "que/that" is observed
are RCs with the grammatical role o f Circumstantial Complement (252 cases). In (168),
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
168
I illustrate the Absence o f a preposition when the relativized position is Circumstantial
Complement:
(168) En un tiempo [(en) que mi sinora no trabajaba....] (Lope Blanch, El Espanol 124)
'At a time [in which my wife did not work....]
In (168), a circumstance of time is expressed by the RC, and the preposition
'en/in' is not used by the speaker with the relative pronoun.
Real Academia Espanola does not consider the Absence of a preposition before
the relativizer "donde/where" a nonstandard use as it does in the case of prepositions
before the relativizer "que/that." Consider (169), in which the relativizer "donde/where"
is preceded by a preposition:
(169) Ahi es-digamos- un seminario [en donde se recibe la formation general para el
sacerdocio, particularmente en el aspecto espiritual....] (Lope Blanch, El Habla de
la Ciudad 56).
'A seminary, [where priests obtain their education, particularly in spiritual
matters....]
In (169), it is also possible to omit the preposition "en/in."
Although the absence of a preposition in cases like (169) is not considered
nonstandard, it is possible that the speakers who are aware of the standard norm that
prescribes the use of prepositions with "que/that" also tend to use a preposition before
"donde/where," because the grammatical role of this relativizer within the RC is
Circumstantial Complement.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
169
4.2. Restrictive and Nonrestrictive RCs, Level of education, and Socioeconomic
level.
The first hypothesis examined in this chapter is that the use o f Restrictive and
Nonrestrictive RCs is influenced by factors such as level o f education, and
socioeconomic level. Macaulay’s study in Ayr, Scotland, demonstrated that there were
social differences in the production o f Restrictive and Nonrestrictive RCs.
Restrictive and a Nonrestrictive RCs are not treated as two variants of the same
variable in this investigation because their inherent features are different, as I have
explained before. For this reason, speakers do not have the option of choosing between
a Restrictive or a Nonrestrictive RC in a given context.
For the reasons listed above, crosstabulations were not made to quantify the use
of Restrictive and Nonrestrictive RCs in the different levels of education and
socioeconomic levels. Instead, the procedure used was to count all words in the first five
pages of each transcribed corpus (corpora from speakers belonging to the middle class
and from the lower class), divide the results by five to obtain the average number o f
words per page, and multiply this average by the total number o f pages of the book (each
of the transcribed corpora). This procedure provided an approximate calculation o f the
total number o f words o f the book. Once the approximate total number o f words in
each transcribed corpus was calculated, it was divided by the total number o f
Nonrestrictive RCs and by the total number of Restrictive RCs. The resulting figure was
the total number of words in which a Restrictive or a Nonrestrictive RC was produced.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
170
The two corpora that were used for the approximate calculation explained above
were El Habla de la Ciudad de Mexico, which contains the speech o f middle class
speakers, and El Habla Popular de la Ciudad de Mexico, which contains the speech of
lower class speakers. Any material not shared by the these two corpora (Conferences
and Monologue, and secret recordings) was not considered in the calculation process.
Two group interviews in which one o f the two speakers was noncollege educated
were not considered in the calculation o f the approximate number o f words of the college
educated corpus.
Because the corpora used in the present investigation was not computerized,
when calculating the approximate number of words o f each transcribed corpus, the
words included in the interviewers' speech were also included, although they were not
considered in the codification of Nonrestrictive and Restrictive RCs.
The results revealed that speakers with a higher level o f education (college
education2 4 ) tend to produce Nonrestrictive RCs more frequently than speakers with a
lower level of education (no college education2 5 ).
Table 19 shows the results of the quantification o f Restrictive and Nonrestrictive
RCs among speakers with college education and speakers with no college education:2 5
2 4 College education includes speakers with any level of college education.
^No college education includes speakers with any level of primary and/or
secondary education, technical school, and speakers without a formal education.
“ Since most of the speakers from the corpus of the middle class (Lope
Blanch, El Habla de la Ciudadl were also college educated, the speakers that were
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
171
Table 19
Frequency of production of Restrictive and Nonrestrictive Relative Clauses among
monolingual speakers with College and No College education
Relative Clause Level of Education
College Education No college education
Restrictive 1 Restrictive RC
every 141 words
1 Restrictive RC
every 196 words
Nonrestrictive 1 Nonrestrictive RC
every 347 words
1 Nonrestrictive RC
every 1248 words
Total 1 RC every 100 words 1 RC every 170 words
Table 19 reveals that college educated speakers tend to produce Restrictive and
Nonrestrictive RCs more frequently than speakers who are not college educated. The
difference between college and no college educated speakers is more clear in their use of
Nonrestrictive RCs (1 Nonrestrictive RC every 347 words among college educated
speakers and 1 Nonrestrictive RC every 1248 words among no college educated
speakers). Table 19 also shows that when combining Restrictive and Nonrestrictive RCs,
college educated speakers tend to produce 1 RC every 100 words, whereas speakers
with no college education produce 1 RC every 170 words.
As Table 20 shows below, middle class speakers tend to produce Restrictive and
Nonrestrictive RCs more frequently than speakers from the lower class:
noncollege educated were not considered in this quantification.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
172
Table 20
Frequency of production of Restrictive and Nonrestrictive Relative Clauses among
monolingual speakers from the middle and lower class.
Relative Clause Socioeconomic level
Middle class Lower class
Restrictive 1 Restrictive RC
every 160 words
1 Restrictive RC
every 196 words
Nonrestrictive 1 Nonrestrictive RC
every 407 words
1 Nonrestrictive RC
every 1248 words
Total 1 RC every 115 words 1 RC every 170 words
The results for the college educated speakers and the results for the middle class
speakers in Tables 19 and 20 respectively differ because a few speakers from the middle
class were not college educated. Middle class speakers exhibit a greater frequency of
Restrictive and Nonrestrictive RCs, and a higher frequency of RCs (Restrictive and
Nonrestrictive RCs combined).
The strongest difference between the two socioeconomic groups in Table 20 is
observable in terms of the production of Nonrestrictive RCs. This result is similar to
Macaulay's observation in Ayr, Scotland, that middle class speakers tend to produce
Nonrestrictive RCs more frequently than speakers from the lower class.
These results suggest that there is a variation o f communicative styles among the
different educational and socioeconomic groups in terms of the speakers’ production of
Nonrestrictive RCs. The difference between the two levels o f education and the two
socioeconomic levels is very clear in terms of the speakers' use o f Nonrestrictive RCs.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
173
Speakers with college education and speakers from the middle class may have a
communication style (at least in terms of their use o f RCs) that tends to use
Nonrestrictive RCs more frequently, as compared to the other groups of speakers, to
achieve the function o f evaluation or commentary in their speech.
The differential access that speakers from different socioeconomic groups have to
the range of oral and written registers has been observed in earlier works (Finegan and
Biber 337-38). Because the written registers have not been analyzed in this work, I can
only speculate that college educated speakers and middle class speakers typically have
more access to the written language, which may have an influence on their higher
frequencies of use of Nonrestrictive RCs, if indeed Nonrestrictive RCs are more frequent
in the written language.
4.3. Presence/Absence of a preposition and Level of Education, Socioeconomic
Level, and Register
This section will examine the hypothesis that the variable Presence/Absence o f a
preposition with a relativizer is sensitive to social variables such as level of education,
socioeconomic level and Register. In the present study, it is assumed that the presence of
a preposition with 'que/that' rather than its absence reflects the standard Spanish norm.
I expect that the Presence of a preposition will be favored by these factors. I also
expect that Conferences and the Monologue will exhibit the highest frequencies of
Presence of a preposition with a relativizer when compared to One-to-one interviews and
Group interviews. Under this hypothesis, speakers should tend to use the standard
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
174
variant (Presence of a preposition) when they are in communicative situations that
involve academic topics, which is the case in Conferences and in the Monologue. In
Tables 21 and 22 I establish correlations of the variable Presence/Absence of a
preposition with "que/that" and "donde/where," respectively, by level of education
among the monolingual speakers:
Table 21
Presence/Absence of a Preposition with the relativizer "que/that" by Level of
Education (Monolingual speakers)
No college
education
College
Education
Total
N % N % N %
Presence of
Prep.
20 16% 76 55% 96 36%
Absence o f
Prep.
109 85% 62 45% 171 64%
Total 129 100% 138 100% 267 100%
p< 000
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
175
Table 222 7
Presence/Absence of a Preposition with the relativizer "donde/where” by Level of
Education (Monolingual speakers)
No college
education
College
education
Total
N % N % N %
Presence of 15 20% 35 49% 50 35%
a Prep.
Absence of
a Prep.
59 80% 36 51% 95 66%
Total 74 100% 71 100% 145 100%
p<000
In Table 21, the Presence of a preposition with "que/that" reaches 55% among
college educated speakers, but it only reaches 16% among the speakers with no college
education. Similarly, Table 22 shows that college educated speakers exhibit Presence of
a preposition with "donde/where" in 49% of the cases, while noncollege educated
speakers show a much lower percentage of Presence of a preposition (20%).
2 7 As can be noticed in Tables 20 and 21, the total number of cases with the
relativizer 'que/that' (267 occurrences) outnumbers the cases with the relativizer
'donde/where' (145 cases). This may be due to the fact that while the relativizer
'donde/where' can only refer to circumstances of place, the relativizer 'que/that'
may be used for other circumstances (time, mode, means, instrument or cause).
Additionally, the alternation of Presence/Absence o f a preposition with 'que/that'
also occur with other syntactic roles (besides Circumstantial Complement) in the
data, such as Direct Object, Object of Preposition, and Indirect Object.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
176
The correlation of the variable Presence/Absence o f a Preposition with the
variable Socioeconomic level reveals a higher frequency of Presence of a preposition
among the middle class speakers, as shown in Table 23:
Table 23
Presence/Absence of a Preposition with the relativizer "que/that" by
Socioeconomic Level (monolingual speakers)
Lower Class Middle Class Total
N % N % N %
Presence o f Prep. 18 14% 80 49% 98 33%
Absence of Prep
111 86% 85 52% 196 67%
Total 129 100% 165 100% 294 100%
p<000
According to Table 23, middle class speakers favor the Presence of a Preposition
in 49% of the cases, whereas lower class speakers only favor it in 14% of the cases.
Notice that when Tables 21 and 23 are compared, the percentage of Presence of a
Preposition is lower among the Middle class speakers in Table 23 (49%), than the one
observed among college educated speakers in Table 21 (55%).
The lower frequency of the standard variant among the middle class speakers may
be due to the fact that some middle class speakers are not college educated. These
speakers' lower level o f education may cause the lower percentage of Presence of a
preposition among the middle class speakers in Table 23, in comparison to the frequency
of Presence of a Preposition among college educated speakers in Table 21.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
177
The correlation with "donde/where" showed a higher frequency of Presence of a
preposition among middle Class speakers. This is shown in Table 24:
Table 24
Presence versus Absence of a Preposition with the relativizer ’’donde/where" by
Socioeconomic Level (monolingual speakers)
Socioeconomic Level Total
Lower Class Middle Class
N % N % N %
Presence 14 20% 38 45% 52 34%
Absence 56 80% 47 55% 103 66%
Total 70 100% 85 100% 155 100%
p<001
The middle class speakers showed Presence of a preposition in 45% of the cases,
whereas the lower class speakers exhibited a lower frequency o f this variant (20%).
The register that exhibits the highest frequency of the standard variant is
Conferences (academic lectures), shown in Tables 25 and 26. In Table 25 I establish the
correlation of Presence/Absence o f a preposition with the relativizer "que/that" by
Register among the Middle class speakers. Table 26 shows this same correlation but
with the relativizer "donde/where":2 8
^ o r the correlations with Register I included the monolingual and the
bilingual speakers in order to count with a sufficient amount o f cases.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
178
Table 252 9
Presence/Absence of a Preposition with the relativizer "que/that" by Type of
Register among the Middle Class speakers
One to One
Interview
Group
Interview
Conference Total
N % N % N % N %
Presence 37 27% 20 36% 15 54% 72 33%
Absence 98 73% 35 64% 13 46% 146 67%
Total 135 100% 55 100% 28 100% 218 100%
p<023
Table 26
Presence/Absence of a Preposition with the relativizer "donde/where" by Type of
Register among the Middle Class speakers
One to One
Interview Group
Interview
Conference Total
N % N % N % N %
Presence 21 22% 11 44% 9 53% 41 30%
Absence 73 78% 14 56% 8 47% 95 70%
Total 94 100% 25 100% 17 100% 136 100%
p<010
In Table 25, it is not surprising that the standard variant (Presence of a
preposition) reaches its highest frequency (54%) in conferences. This is also observable
“Tables 24 and 25 do not include Monologue, because initial correlations
showed empty cells. The 11 cases in Monologue with ‘que/that’ occurred with
Presence o f a preposition, whereas the 5 cases of Monologue with ‘donde/where’
occurred with Absence of a preposition.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
in Table 26, in which conferences show a similar percentage (53%) for Presence of a
Preposition with the relativizer 'donde/where.'
When establishing a correlation of the variable Presence/Absence of a
preposition by Register among College educated speakers, One to one interviews exhibit
the highest frequency of the standard variant, as shown in Table 27:
Table 27
Presence/Absence of a preposition with the relativizer "que/that" by Register
among College educated speakers
One to One
Interview
Group
Interview Conference Total
N % N % N % N %
Presence 30 60% 20 41% 15 54% 65 51%
Absence 20 40% 29 59% 13 46% 62 49%
Total 50 100% 49 100% 28 100% 127 100%
p< 155
Conferences in Table 27 did not show the highest frequency of Presence of a
preposition (54%), while one-to-one interviews exhibited the highest percentage of the
elaborated variant (60%). However, this correlation did not result in statistically
significant findings.
The results of Table 27 are similar to the frequencies obtained after establishing a
correlation with the relativizer "donde/where" among college educated speakers. This
correlation did not result in significant statistics, either (p<817), but showed that
of the copyright owner. Further reproduction prohibited without permission
180
Conferences exhibited 53% o f cases of Presence of a preposition, and One-to-one
interviews exhibited 57% o f Presence o f a preposition.
The statistically significant results shown above (Tables 25 and 26) suggest that
Middle class speakers tend to consider the Presence o f a preposition as a standard
variant, since it is more frequently associated with Conferences.
Because the Conferences in the present investigation were about academic topics,
it is not surprising that the Conferences exhibit the highest frequencies o f Presence of a
preposition among the middle class speakers (Tables 25 and 26), when compared to the
other registers analyzed. Although these Conferences were not simply readings o f texts,
typically the speech in this type o f register is closer to written language in comparison to
One-to-one interviews and Group interviews.
4.4. Presence/Absence of a preposition and Monolingual/Bilingual
Bilingual speakers do not tend to favor the Presence of a preposition (the more
standard variant) more than monolingual speakers with no college education. In Table
2 8 ,1 establish a correlation o f the variable Presence/Absence of a preposition with the
relativizer "que/that" and with three differentiated groups of speakers: monolinguals with
no college education, monolinguals with college education, and bilingual speakers. In
Table 29 I establish the same correlation, but with the relativizer '"donde/where."
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
181
Table 28
Presence/Absence of a Preposition with the relativizer "que/that" among
Monolinguals with No college education, Monolinguals with College education, and
Bilinguals
Noncollege
Educated
Monolinguals
College
Educated
Monolinguals Bilinguals
Total
N % N % N % N %
Presence 20 16% 76 55% 4 5% 100 29%
Absence 109 85% 62 45% 79 95% 250 71%
Total 129 100% 138 100% 83 100% 350 100%
p<000
Table 29
Presence/Absence of a Preposition with the relativizer "donde/where" among
Monolinguals with No college education, Monolinguals with College education, and
Bilinguals
Noncollege
Educated
Monolinguals
College
Educated
Monolinguals
Bilinguals Total
N % N % N % N %
Presence 15 20% 35 49% 4 6% 54 26%
Absence 59 80% 36 51% 60 94% 155 74%
Total 74 100% 71 100% 64 100% 209 100%
p< 000
As predicted, monolingual speakers with college education favor the standard
variant the most (55% in Table 28 and 49% in Table 29).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
182
The bilingual speakers in Tables 28 and 29 exhibit lower frequencies of Presence
o f preposition than the monolingual speakers with college and no college education. The
bilingual speakers show 5% of presence o f a preposition in Table 28 and 6% of this
variant in Table 29.
With Presence/Absence of a preposition with "que/that," the Presence of a
Preposition is the standard norm in Spanish. This may be the reason why bilingual
speakers are less aware than the monolingual speakers (with college and no college
education) of the Spanish normative use of prepositions when the relativizer "que/that"
has the grammatical role of Circumstantial Complement.
The English and Spanish languages have parallel structures to express the
Circumstantial Complement within the RC, as shown in (170) and (171):
(170) The house [where I live] is pretty
E a casa [(en) donde vivo] es bonita1
(171) The months [in which we worked on that project] were very busy
Eos meses [(en) que trabajamos en ese proyecto] fueron muy ocupados'
In English, a preposition is not used before the relativizer "where/donde,"
whereas the use of the preposition before "donde/where" is possible in Spanish, as
observed in (170). Similarly, Spanish allows the variable use of a preposition with
"que/that" in (171). whereas this variation does not occur in English; that is, the Absence
o f a preposition does not occur in examples like (171) in English.
The fact that prepositions are not used before "where/donde" in English may
explain the 6% of Presence of a preposition among bilingual speakers in Table 29. If the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
183
bilingual speakers never use a preposition with ‘where/donde’ in situations like (170) in
English, this may influence their infrequent use of a preposition with "donde/where"
when they use Spanish. However, if the bilingual speakers do not omit prepositions as in
(171) in English, it could be predicted that when they use Spanish their use of
prepositions with "que/that" would be higher than the 5% observed for Presence of a
preposition in Table 28.
The explanation for the low frequencies reached by Presence o f a preposition
with a relativizer "que/that" among the bilinguals, in comparison to monolinguals (college
and noncollege educated), may be that the bilingual speakers were not exposed to
standard Spanish use during their formal education, which recommends the use of a
preposition with the relativizer "que/that" in examples like (170). Even the monolingual
noncollege educated speakers have had, in many cases, some instruction in the Spanish
language. This could explain that the monolingual noncollege educated speakers tended
to use prepositions with the relativizer "que/that" more frequently than the bilingual
speakers.
As I have explained earlier, the low percentage of use of prepositions with the
relativizer "where/donde" in Spanish may be due to the fact that the bilingual speakers
never use a preposition with the relativizer "where/donde" in English.
4.5. Summary and Conclusions
The correlation o f the use o f Restrictive and Nonrestrictive RCs with social
factors shows that the use of Nonrestrictive RCs was favored by speakers with a higher
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
184
level of education and higher socioeconomic level. In fact, the speakers with college
education and the speakers from the middle class exhibit higher frequencies o f use of
Restrictive RCs, and Restrictive and Nonrestrictive RCs combined. However, as seen in
Tables 19 and 20, the strongest difference between college and noncollege educated
speakers and between middle and lower class speakers is observed with regard to the use
of Nonrestrictive RCs.
The clear social differentiation that is observed in the use of Nonrestrictive RCs
suggests that the speakers from the middle class and the speakers with college education
may have a style of communication that tends to use evaluative structures, at least in
terms of their use of RCs. Because Nonrestrictive RCs have the inherent quality of being
structures that typically convey a comment about an already specified antecedent, these
structures are an appropriate device to achieve an evaluative role with respect to their
antecedents.
The analysis of the variable Presence/Absence of a preposition revealed that the
more standard variant (i.e., the Presence o f a preposition) was favored by those speakers
with a higher level o f education and a higher socioeconomic level. These results may be
due to the greater amount of exposure that college educated speakers and middle class
speakers typically have to the written language, which tends to favor the use o f standard
variants.
The bilingual speakers did not produce higher frequencies of this variant than the
noncollege educated monolingual speakers. This may be because monolingual speakers
are more familiar than bilingual speakers with the prescriptive use of prepositions when a
with permission of the copyright owner. Further reproduction prohibited without permission.
185
relativizer "que/that" has the role o f Circumstantial Complement. The bilingual speakers
also tended to produce the lowest frequency o f Presence o f a Preposition with the
relativizer "donde/where." The fact that "where/donde" in English is not used with
prepositions when it has the role of Circumstantial Complement may have an influence on
the low frequency of use of a preposition with the relativizer "donde/where" among
bilingual speakers, in comparison with monolingual speakers.
A result that was consistent in the analysis o f the variable Presence/Absence of a
preposition in various Registers was that Conferences are associated with higher
frequencies of Presence of a preposition among the middle class speakers. The fact that
the Conferences in the present investigation involved academic topics may explain that
the standard variant (Presence o f a preposition) is more frequent in this type o f Register.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 5
186
CONCLUSIONS
In this dissertation, I analyze correlations between the grammar of Spanish
relative clauses and the semantic/pragmatic factors that are present in the interaction.
I also examine the correlation of social factors with the variable use of Restrictive and
Nonrestrictive RCs and the influence of social and situational factors in the variable use
o f a preposition with a relativizer.
5.1. Semantic/Pragmatic factors
5.1.1. Humanness
My first hypothesis is that Humanness has an influence on patterns of
relativization for Human and Nonhuman antecedents in Spanish RCs. The results of this
investigation have shown the importance of the factor Humanness in the distribution of
syntactic roles in RCs, and in their antecedents. Human antecedents tend to be
grammatical Subjects, and they also tend to be relativized as Subjects. The tendency for
Humans to refer to other Humans (Fox and Thompson 304) and the tendency for Human
referents to be coded as Subjects in Spanish (Silva-Corvalan, "On the Interaction" 119)
may explain that speakers have developed a strategy to relativize Subjects.
Nonhuman antecedents are more frequently associated with the roles of
Circumstantial Complement and Direct Object. Furthermore, although Object relatives
do not outnumber Subject relatives with Nonhuman antecedents, the frequency o f Object
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
187
relatives increases with Nonhuman antecedents, in comparison to Human antecedents.
These observations supported the general tendency observed in Spanish for Nonhuman
referents to be coded as Direct Objects (Silva-Corvalan, "On the Interaction" 119). The
fact that Subject relatives do not outnumber Object relatives with Nonhuman antecedents
may be due to the fact that Nonhumans tend to be referred to in terms o f their owners or
manipulators. Indeed, it has been noticed “the major way that nonhuman referents are
made relevant is in terms o f the humans who own them, use them, and manipulate them.”
(Du Bois 269-70, cited by Fox and Thompson 304)
Object relatives provide a mechanism by which an antecedent can be related to a
referent (usually a Subject pronoun that is a Subject within the RC). In a typical Object
relative, a Subject pronoun is the Subject of the RC, whereas the coreferent of the
antecedent is the Direct Object within the RC. This Subject-Direct Object relationship
in the RC tends to be associated with relationships of ownership, use, or manipulation
between a Human and a Nonhuman referent. As I explained in Chapter 3, my results
supported Fox and Thompson's (304) observation that this syntactic configuration of
Object relatives makes them more appropriate than Subject relatives for the task of
Anchoring Nonhuman Subject antecedents in their study.
However, Object relatives in my investigation are frequently found performing
Anchoring of Human and Nonhuman antecedents with various syntactic roles.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
188
5.1.2. The factors Humanness, Grounding, and Type of Combination
My second hypothesis is that the speakers’ selection o f different ways of
grounding referents in the discourse (by Anchoring, by Main-Clause grounding, or by
Proposition-Linking grounding) is influenced by the Humanness of the 11 TCs that are
studied.
I have shown that Object relatives are more frequently associated with Anchoring
than Subject Relatives. It was observed that for every grammatical role in the antecedent
(Subject, Direct Object, Object o f Preposition, and Predicate Nominal), Object relatives
correlate with higher frequencies of Anchoring (as compared to other types of
grounding) than Subject relatives. This occurs with Human and Nonhuman antecedents.
Fox and Thompson (303) have found in English that Object relatives
outnumbered Subject relatives when the antecedents were Nonhuman because o f the
speakers' need to ground Nonhuman antecedents by Anchoring. My data does not show
a preference for Object relatives over Subject relatives with Nonhuman antecedents, but
it does reveal that Anchoring is more frequent with Object relatives than with Subject
relatives.
Object relatives are better suited than Subject relatives for the task of Anchoring
an antecedent because they typically contain a Human Evoked Subject Pronoun (which
serves as an anchor) whose referent is different from the referent of the antecedent.
My results suggest that Anchoring is a mechanism by which Human Evoked
referents ground Human and Nonhuman entities in the discourse. When Nonhuman
antecedents are being anchored by Humans, the element that receives the Grounding and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
189
the element that provides the Grounding usually establish a relationship o f ownership and
manipulation.
Fox and Thompson's observation that in oral interaction the major way that
Nonhuman referents are made relevant is in terms of the Humans who own them, use
them, and manipulate them (Fox and Thompson 304) is confirmed in my analysis of
Anchoring by Object relatives. In Anchoring of Humans, the relationship is not one of
manipulation, usage, or ownership, but typically involves a relationship between two
Human referents in which one referent is grounded by means o f its relationship to the
other referent.
Although Humans do not typically establish relationships of ownership with the
Human referents that they anchor, I found that Human referents that need Grounding
also need to be related to Human (Subject) Evoked referents to obtain relevance. The
second type of Grounding studied in this dissertation, Main-Clause grounding, is also
associated with the syntactic role of Direct Object for Human and Nonhuman
antecedents, but within the main clause rather than within the RC, as with Anchoring.
When an antecedent has the role of Direct Object, typically a Subject has already been
uttered. This Subject frequently grounds the antecedent of the RC (the referent of the
Direct Object) by Main-Clause grounding. The typical SVO word order in Spanish may
have an influence on the preference for the role of Direct Object over Subject in the
antecedents when there is Main-Clause grounding. The association between antecedents
with the role o f Direct Object and Main-Clause grounding has been already noted by Fox
and Thompson (305) for Nonhuman antecedents in English.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
190
The third type of Grounding studied in this work, Proposition- Linking
grounding, exhibits different patterns than those observed with Anchoring and Main-
Clause grounding. Proposition- Linking was favored by TCs that did not have
antecedents with the grammatical role of Direct Object, that is, 'pn-s,' 's-s,' ’ex-s,1 and
'pn-do.' Notice that with the exception of'pn-do,' all o f these TCs lack the role of Direct
Object, whether in the antecedent or in the RC. When an antecedent is grounded by
Proposition- Linking, it does not need to be grounded by Humans in the main clause
(Main-Clause grounding) or by Humans in the RC (Anchoring). As I have pointed out
above, the referent of a Direct object in the main clause tends to be grounded by the
referent of the Subject in the main clause (Main Clause grounding). When Grounding is
done by Anchoring, a Human Subject pronoun is frequently the element that is related to
the antecedent of the RC in order to provide the Grounding.
If Proposition- Linking grounding occurs, the grammatical role favored in the
antecedent does not tend to be a Direct Object because there is no need for Human
Subject referents to ground these referents which have already been made relevant by
Proposition- Linking grounding, due to their link to the preceding context.
The patterns discussed above are exhibited by Human and Nonhuman antecedents. The
grammatical choices that the speakers make in terms of the TCs they use in relativization
influences the ways they select to ground referents in the discourse. My observation, that
it is not the factor Humanness but the grammatical role that favors different ways of
grounding a referent, advances our knowledge of the interaction o f semantic/pragmatic
factors beyond what was detailed by Fox and Thompson, who found discourse
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
191
explanations for the occurrence o f the TCs 's-do' and 'do-s' for Nonhuman antecedents,
and for the TC 'ex-s' with Human antecedents. My study includes a wider range of
syntactic possibilities observed in relativization in Spanish; and it also considers Human
and Nonhuman antecedents. The analysis o f Human and Nonhuman antecedents has
demonstrated that the interaction of different syntactic roles involved in relativization are
directly associated with different modes of establishing the relevance of the referents o f
the RCs and their antecedents.
5.1.3. The factors Functional Role of the RC and Type of Combination
My third hypothesis is that the TCs that I analyze (with Human and Nonhuman
antecedents) determine various functional roles achieved by the RCs, and that Grounding
has an influence on these tendencies. Fox and Thompson (301-2) have analyzed two
different functional roles that RCs may have with respect to their antecedent. These
functions are Characterization and Identifying an Evoked antecedent (“helping to
identify” a Given Head NP, in their terminology). Fox and Thompson have noted in their
data that Identification and Grounding by Anchoring are usually performed by Object
relatives with Nonhuman Subject antecedents (303). These authors have also observed
the association of grounding by Proposition Linking and the use o f Subject relatives with
Existential Human antecedents (309). With Main-Clause grounding of Nonhuman Direct
Object antecedents, Fox and Thompson (305-07) explained that Subject relatives tend to
perform Characterizations.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
192
The function o f Giving New Information about an antecedent (which could be
Evoked, Brand-New or Inferrable) is the most prevalent o f the functional roles studied in
this dissertation, which are Characterization, Giving New Information, Identification, and
Reaffirming the Existence o f an antecedent. Subject, Object, and Circumstantial
Complement relatives3 0 tend to favor Giving New Information about an antecedent, but a
closer examination of the results reveals that there are differences between Human and
Nonhuman antecedents in terms o f the second most favored function by Subject and
Object relatives. Among Human antecedents, the second most preferred function for
Subject relatives is Characterization, but Object relatives exhibit a second preference for
Identification. Identifications are predetermined by the Evoked Information Status of the
antecedent and by the Evoked informational value within the RC.
Among Nonhuman antecedents, Subject relatives also favor Characterizations as their
second preference. On the other hand, Object relatives exhibit a second preference for
Reaffirming the Existence o f an antecedent. The function o f Reaffirming the Existence o f
an antecedent presupposes either Evoked or Inferrable antecedents, but the high
percentage o f Object relatives performing this function indicates that Nonhuman referents
may tend to be reaffirmed in the discourse by being related with Human referents. Object
relatives provide an appropriate device for Human referents (typically a Subject pronoun
^Notice, however, that there were only 3 cases o f Circumstantial
Complement relatives with Human antecedents, as was explained in Chapter 3.
For this reason, and because there were empty cells, these cases were excluded
from the correlation. As was noticed in Chapter 3, 2 o f the 3 cases of this type of
RC occurred in the function of Giving New Information.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
193
within the RC) to establish a Subject-Direct Object relationship in which the Direct
Object within the RC is the coreferent o f the antecedent.
The correlation of the different TCs with Functional Role of the RC by Anchoring
and Main Clause grounding reveals that the most preferred function for Subject and
Object relatives (Human and Nonhuman antecedents) is Giving New Information about a
antecedent. When there is a clear second preference in terms o f Functional role with
Anchoring or Main Clause grounding (Human or Nonhuman antecedents), it tends to be
Characterization for Subject relatives, and either Identification or Reaffirming the
Existence o f an antecedent for Object relatives.
With Proposition- Linking grounding of Human antecedent, some differences are
observed with respect to Anchoring and Main-Clause grounding. Among Human
antecedents, there is an overall increase of the frequency of Characterizations. In fact,
with Human antecedents grounded by Proposition-Linking grounding, Characterization
exhibits a significant increase. As shown in Table 13, Characterizations occurs in 38% of
the cases, whereas Giving New Information has a frequency o f 35% of the total number
of cases. The TCs that favor or that have high frequencies o f Characterizations are
Subject relatives. These results support Fox and Thompson's (309) observation that
Subject relatives are frequent with Proposition Linking grounding because there is no
need for Human antecedents to be related to other Human referents that ground them (as
with Object relatives). Thus, Subject relatives with Proposition Linking grounding tend
to refer to the antecedents' own features and activities. This favors Characterizations.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
194
With Nonhuman antecedents, the correlation with the different TCs shows that
when the antecedents are grounded by Proposition-Linking, the TCs with Subject
relatives reach very high frequencies o f Characterizations, and in some cases the
frequencies reached by Characterizations with the TCs that have Subject relatives
outnumber Giving New Information with Subject relatives. However, the most favored
function is Giving New Information, and Reaffirming the Existence o f an antecedent is
the second most frequent function, favored by TCs with Object relatives. This tendency
for Object relatives to be as a device used by the speakers to reaffirm the existence of
antecedents of RCs may be due to the fact that Object relatives (unlike Subject relatives)
may be a mechanism by which the speakers can validate Nonhuman antecedents by
relating them with Human referents, which are typically the Subject pronouns within the
Object relative. This validation o f Nonhuman referents by Humans in the discourse can
explain the increase of the function o f Reaffirming the Existence of an antecedent among
Nonhuman antecedents.
Because Reaffirming the Existence of an antecedent involves Evoked or
Inferrable antecedents that are validated in the conversation typically by Humans, and
because the information contained in the RC when this function occurs is Inferrable, the
association of this functional role with Proposition-Linking grounding is not surprising,
given that this type of Grounding consists of antecedents of RCs being grounded by their
relation to earlier contexts in the interaction, which tends to favor Inferrable or Evoked
antecedents, and, very infrequently, Brand-new antecedents.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
195
The function of Identification o f an antecedent is predetermined by the Evoked
Information Status of the antecedent and the Evoked informational value within the RC.
The high percentage o f occurrence o f the TC 'pn-do' with Identification is due to the
high percentage of Evoked antecedents and Evoked informational value within the RC
with this TC.
The TC 'pn-do,' when there is Identification of a Nonhuman antecedent, is
frequent with Proposition- Linking grounding because this type of Grounding favors
Evoked or Inferrable antecedents. Grounding by Proposition- Linking implies an
antecedent's link to a preceding context.
Identification and Reaffirming the Existence o f a Nonhuman antecedent favors Object
relatives over Subject relatives when there was Proposition Linking. With Proposition
Linking, the antecedent of a RC is made relevant by its link to the preceding context.
Although the antecedent has already obtained relevance by its link to the preceding
context, its identification or validation in the discourse is achieved by its link to Human
Evoked Subjects that are different from such an antecedent, and this favors Object
relatives rather than Subject relatives.
My examination of patterns o f relativization in this work has departed from Fox
and Thompson's findings in English conversation, but has also found some similar results.
However, by doing a more extensive analysis o f the semantic/pragmatic factors involved
in relativization, and by incorporating a wider range o f syntactic possibilities, I have
shown more general patterns in terms of the interaction o f Humanness, Grounding,
Information Status, and Functional role an RC with the grammatical choices made by the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
196
speakers in the process of relativization. The coincidences between my findings and Fox
and Thompson's suggest that relativization may be influenced by the semantic/pragmatic
factors, studied here, in other languages.
The present study also provides an approach different from the typological
perspective proposed by Keenan and Comrie and Comrie and Keenan. The perspective
in my study is that the variation observed in relativization is motivated by pragmatic
rather them cognitive factors. In this dissertation I have shown that the speakers'
development of different strategies to relativize Human and Nonhuman referents is
influenced by factors that are present in the interaction. My investigation has explained
the grammar of RCs from a functional perspective, which considers language a system of
communication wherein the grammar of RCs is shaped by the speakers' communicative
pressures in the discourse.
5.2. Social and discourse variables
In Chapter 4 ,1 examined the correlation of the use of Restrictive and
Nonrestrictive RCs with level of education and socioeconomic level. I also analyzed the
correlation of the variable Presence/Absence of a preposition with level of education,
socioeconomic level, register, and the variable Bilingual/Monolingual.
5.2.1. Restrictive and Nonrestrictive RCs and their correlation with social factors
The main distinctive feature o f Nonrestrictive RCs is their evaluative feature,
following Labov's definition of evaluation (Language 366-70), since they tend to provide
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
197
a commentary about antecedents whose reference has already been specified in the
discourse.
My hypothesis is that level of education and socioeconomic level have an
influence on the variation observed in the production of Restrictive and Nonrestrictive
RCs. The results o f the quantification o f the use of Restrictive and Nonrestrictive with
level of education and socioeconomic level show that college educated speakers and
speakers who belong to the middle class tend to favor higher frequencies of use of
Nonrestrictive RCs than noncollege educated speakers and lower class speakers. These
results support Macaulay's findings in Scots’ RCs: the higher frequency of Nonrestrictive
RCs among middle-class speakers. These results suggest the existence o f different styles
of communication among the different levels of education and among the different
socioeconomic levels, at least in terms of the speakers' use of RCs. The college educated
speakers and the middle class speakers may have a style of communication in their
relativization that exhibits a more frequent use of evaluative structures, i.e.,
Nonrestrictive RCs.
The results in Chapter 4 also show that the use of Restrictive RCs and of
Restrictive and Nonrestrictive RCs combined is more frequent among college educated
speakers and among middle class speakers. However, the social differentiation in the use
of RCs is more clear in terms of the speakers’ use of Nonrestrictive RCs.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
198
5.2.2. The analysis of the correlation o f Presence/Absence of a preposition with
Social variables and with register
My hypothesis in this section was that the variable Presence/Absence o f a
preposition may be sensitive to the variables Level o f education, Socioeconomic level,
and Register.
In this study, it is assumed that Presence o f a preposition with a relativizer is the
standard variant. The correlations of the Presence/Absence of a preposition with the
variable Level of Education showed that the non-college educated speakers exhibited
considerably higher frequencies o f Absence o f preposition (with the relativizers
"que/that" and "donde/where") than speakers with a college education.
The correlation with socioeconomic level showed that lower class speakers
tended to exhibit higher percentages of Absence o f the preposition than the middle Class
speakers for the two types o f relativizer ("que/that" and "donde/where").
The statistically significant results of the correlation o f Presence/Absence o f a preposition
by Register showed that middle class speakers associated Nonrestrictive RCs with
Conferences. The high frequency o f the standard variant (Presence of a Preposition) in
Conferences may be influenced by the fact that the Conferences of the present study are
academic lectures, which belong to the speakers' professional areas. It is not surprising
that the standard variant (Presence o f a preposition) occurs more frequently at academic
Conferences than in One-to-one or Group interviews.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5.2.3. The analysis of the correlation of Presence/Absence of a Preposition with the
variable Bilingual/Monolingual
When noncollege educated monolinguais are compared to bilingual speakers in
regard to the variable Presence/Absence of a preposition with a relativizer, bilingual
speakers showed a lower frequency o f the standard variant (Presence of a preposition)
than the noncollege educated monolinguais.
In the case of this variable, the level o f education that the bilingual speakers have in
Spanish may cause this result. Considering that the Spanish normative grammar
recommends the use of prepositions whenever the relativizer 'que' has the role of
Circumstantial Complement, the fact that the lower class speakers have a higher degree
o f exposure to Spanish when compared to bilinguals may influence the bilingual speakers'
lower frequency of use o f the normative structure, that is, the Presence of a preposition.
With the relativizer "donde/where," the lower percentage of use of prepositions
by the bilinguals may be due to the fact that in English the relativizer "where" is never
preceded by a preposition when it has the role o f Circumstantial Complement.
5.3. Final remarks
RCs have been studied from several different perspectives, for instance, from a
generative approach (Rivero), or from a typological perspective (Keenan and Comrie;
Comrie and Keenan; Comrie).
This work has shown that patterns of distribution o f syntactic roles of RCs and
their antecedents are statistically associated with semantic / pragmatic factors such as
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 0 0
Humanness, Grounding, Information status, and Functional role of the RC with respect
to its antecedent.
The speaker's various grammatical choices in the process of relativization favor
different ways in which referents are made relevant in the discourse, along with specific
functions that the RCs have with respect to their antecedents.
The present study has shown that the use of Nonrestrictive RCs may be
influenced by the speakers' need to convey commentary about an already specified
antecedent. The higher frequencies of use of Nonrestrictive RCs among speakers with
college education and among middle class speakers suggest that these groups of speakers
may have a style of communication that tends to be more evaluative in terms of their use
of RCs.
The analysis of the variable Presence/Absence of preposition with a relativizer has
shown that the speakers who typically have greater exposure to standard and/or written
language tend to exhibit a higher frequency of use of the standard variable (Presence of a
preposition), when compared to noncollege educated speakers and with lower class
speakers.
Because the Presence of a preposition with a relativizer in some contexts is a
normative mandate in Spanish, the more limited exposure that bilingual speakers have to
Spanish in comparison to monolingual speakers explains why bilingual speakers produced
even lower percentages o f Presence of a preposition than the monolingual speakers with
non-college education. The academic nature of the Conferences detailed in the present
investigation may explain the fact that this register exhibits the highest frequency of use
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 0 1
Presence o f a preposition in comparison to the other types of registers among the middle
class speakers.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 0 2
Bibliography:
Alarcos Llorach, Emilio. Gramatica de la Lengua Espafiola. Madrid: Espasa Calpe, 1994.
Barriga-Villanueva, Rebeca. "La Production de Oraciones Relativas en Ninos Mexicanos
de Seis Anos." Nueva Revista de Filologia Hispanica. 34 (1985-6): 108-55.
Bello, Andres. Gramatica de la Lengua Castellana Destinada al Uso de los Americanos.
Ed. Ramon Trujillo, Madrid: Arcos, 1988.
Campos, Hector. De la Oration Simple a la Oration Compuesta. Curso Superior de
Gramatica Espafiola. Washington, DC: Georgetown Univ. Press, 1993.
Chafe, Wallace. "The Deployment o f Consciousness in the Production o f a Narrative."
The Pear Stories. Ed. Wallace Chafe. Norwood, NJ: Ablex, 1980. 9-50.
— . "Cognitive Constraints on Information now ." Coherence and Grouding in Discourse.
Ed. Russell Tomlin. Amsterdam: Benjamins, 1987. 21-51.
Comrie, Bernard. Language Universals and Linguistic Typology. Oxford: Basil Blackwell,
1989.
Comrie, Bernard, and Edward Keenan. "Noun Phrase Accessibility Revisited." Language
55 (1979): 649-64.
Du Bois, John. "The Discourse Basis o f Ergativity." Language 63 (1987): 805-55.
Finegan, Edward, and Douglas Biber. "Register and Social Dialect Variation: An
Integrated Approach." Sociolinguistic Perspectives on Register. Ed. Douglas
Biber and Edward Finegan. New York: Oxford University Press, 1994. 315-47.
Fox, Barbara."The Noun Phrase Accessability Hierarchy: Subject Primacy or the
Absolutive Hypothesis?" Language. 63 (1987): 856-70.
Fox, Barbara, and Sandra Thompson. "A Discourse Explanation of the Grammar of
Relative Clauses in English Conversation." Language 66 (1990): 297-316.
Hunt, K.W. Grammatical Structures Written at Three Grade Levels. Champaign: National
Council of Teachers of English, 1965.
Keenan, Edward, and Bernard Comrie. "Noun phrase accessibility and universal
grammar." Linguistic inquiry. 8 (1977): 63-99.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
203
Klein-Andreu, Flora. "Grammar in Style: Spanish Adjective Placement." Discourse
Perspectives on Syntax. Ed. F. Klein-Andreu. New York: Academic, 1983. 143-
79.
Kroch, Anthony S., and Donald M. Hindle. A Quantitative Study of the Syntax of Speech
and Writing. 1982. Final Report to the National Institute of Education. Cited in
Edward Finegan and Douglas Biber. "Register and Social Dialect Variation: An
Integrated Approach." Sociolinguistic Perspectives on Register. Ed. Douglas
Biber and Edward Finegan. New York: Oxford University Press, 1994. 346.
Labov, William. Sociolinguistic Patterns. Philadelphia: University of Pennsylvania, 1972.
— . Language in the Inner City. Philadelphia: University of Pennsylvania, 1972.
Lambrecht, Knud. Information Structure and Sentence Form. Topic. Focus and the
Mental Representation of Discourse Referents. Cambridge: Cambridge University
Press, 1994.
Lehman, Christina. Stress and Sentence Position: Ways of Indicating Discourse
Prominence. Discourse Across Time and Space. Ed. Elinor Ochs Keenan and Tina
L. Bennett. Southern California Occasional Papers in Linguistics 5. University of
Southern California: Department o f Linguistics, 1977. 109-21.
Lindenfeld, Jacqueline. 1969. "The Social Conditioning of Syntactic Variation in French."
Advances in the Sociology of Language. Ed. Joshua A. Fishman. Vol. n. The
Hague: Mouton, 1972. 79-90. Rpt. O f American Anthropologist 71(1969): 890-
98.
Lope Blanch, Juan M., ed. El Habla de la Ciudad de Mexico. Mexico City: Universidad
Nacional Autonoma de Mexico, 1971.
—, ed. El Habla Popular de la Ciudad de Mexico. Mexico City: Universidad Nacional
Autonoma de Mexico, 1976.
—, ed. El Espanol Hablado en el Suroeste de los Estados Unidos. Mexico City:
Universidad Nacional Autonoma de Mexico, 1990.
Macaulay, Ronald. Locating Dialect in Discourse. The Language of Honest Men and
Bonnie Lasses in Ayr. New York: Oxford University Press, 1991.
McCawley, James D. "Relative Clauses". The Syntactic Phenomena of English, Vol. 2.
Chicago: University of Chicago Press, 1988. 417-63.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
204
Norusis, Marija J. SPSS for Windows: Base System User's Guide. Chicago: SPSS Inc,
1993.
Prince, Ellen. Toward a Taxonomy o f Given-New Information. Radical Pragmatics. Ed.
Peter Cole. New York: Academic Press, 1981.
Real Academia Espafiola. Esbozo de una Nueva Gramatica de la Lengua Espafiola.
Madrid: Espasa Calpe, 1977.
Rivero, Maria Luisa. Las Construcciones de Relativo. Madrid: Taurus Universitaria, 1991.
Rosenbaum, Harvey. "Relative Clause Structures in Informal Conversation." Studies
Presented to Emmon Bach bv his Students. Ed. Elisabeth Engdahl and Mark J.
Stein. Amherst: University of Massachussetts, 1979. 165-73.
Silva-Corvalan, Carmen. "On the Interaction of Word Order and Intonation: Some OV
Constructions in Spanish." Discourse Perspectives on Syntax. Ed. Flora Klein-
Andreu. New York: Academic, 1983. 117-40.
— . Sociolinguistica: Teoria v Analisis. Madrid: Alhambra, 1989.
Van den Broeck, Jef. Class Differences in Syntactic Complexity in the Flemish town of
Maaseik. Language in Society 6(1977): 149-81.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
APPENDIX 1
205
1. The different TCs of syntactic roles of RCs and their antecedents
47 different types o f possible combinations were obtained in the coding of the
3345 Restrictive and Nonrestrictive RCs coded in this investigation. The 11 TCs
considered in this study were exemplified in Chapter 2. For this reason, I only exemplify
here the remaining 36 TCs found in the data.
For this investigation I considered the TCs with more frequent occurrences in the
data.
The grammatical roles Circumstantial Complement and Adverbial were considered
Circumstantial Complement, both within the RCs and in their antecedents. The Adverbial
relativizers 'cuando/when' and 'donde/where' refer to time and place, and this is a feature
that these elements share with Circumstantial Complements o f time and place.
Subject-Circumstantial Complement:
(1) ...hasta que llego un momento [en que dije:..] (Lope Blanch, El Habla de la
Ciudad 17).
'...until a moment arrived [in which I said:...]'
Object of Preposition-Existential:
(2) Aqui en Mexico, de las gentes mas inquietas [que habia entonces,...] (Lope Blanch,
El Habla de la Ciudad 70).
Here in Mexico, out o f the most active people [that existed back then,...]'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
206
Existential-Object of Preposition:
(3) ...hay cosas [con las que nace u n o ,...] (Lope Blanch, El Habla de la Ciudad 220)
'...there are things [with which one is born,...]'
Predicate Nominal-Object of Preposition:
(4) ...y eran matematicas hasta cierto punto elevadas. [con las cuales ya mi
razonamiento no llegaba a mas,...] (Lope Blanch, El Habla de la Ciudad 20).
'...and they were advanced mathematics, [with which my reasoning would not go
further,...]'
Direct Object-Circumstantial Complement:
(5) ...con el fin de que los ninos lisiados tuvieran una escuela. un kinder, una
primaria. [en la que se Ies diera la ensenanza del tipo...] (Lope Blanch, El Habla de
la Ciudad 15).
'...with the purpose the crippled children would have a schooL a kindergarten, a
primary school. [in which they would received teaching o f the type...]'
Direct Object-Existential:
(6) ...yo habia visto las politicas [que habian dentro de la escuela de Educacion
Fisica,...] (Lope Blanch, El Habla de la Ciudad 16).
'...I had already seen the politics [that would exist within the Physical Education
school,...]'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
207
Direct O bject-Object o f Preposition:
(7) No encontramos en la naturaleza un solo caso. un solo caso. [en que
observacionalmente podamos decir:...] (Lope Blanch, El Habla de la Ciudad 359).
We do not find in nature a single case, a single case, [in which by observation we
are able to say:...]'
Circumstantial Complement-Direct Object:
(8) ...a mi me dijeron las personas que me conocian, que una buena profesion para mi
seria la de diplomatico o licenciado, por tanto [que les hablaba y les platicaba...]
(Lope Blanch, El Habla de la Ciudad 16).
'...I was told by the persons who knew me, that a good profession for me would be
the one of diplomat or licentiate, because of the many things [that I told them and
talked to them about...]'
Object of preposition-Object of Preposition:
(9) ...mefui conun...conun...conunsenor... [queyo trabajaba...] (LopeBlanch, El
Habla Popular 19).
'...I went with ...with a...with a man... [that I worked with...]1
Predicate Nominal-Existential:
(10) Bueno, es una maquina especial [que...que hay, ...] (Lope Blanch, El Habla
Popular 32).
Well, it is a special machine [that...that exists,...]'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
208
Subject-Object of Preposition:
(11) ...la larga prole [con que me habia obsequiado esta sefiora...] pues...me obligaba...a
muchas.-.necesidades que cubrir,... (Lope Blanch, El Habla de la Ciudad 102).
'...the large family [this lady had provided me with...] well...it would force me...to
provide for several necessities,...1
Existential-Circumstantial Complement:
(12) Hay veces [que se me presenta alguna otra cosa importante,...] (Lope Blanch, El
Habla de la Ciudad 93).
'There are times [in which some other important thing occurs,...]'
Existential-Direct Object:
(13) Hay muchas cosas [que uno comprende...] (Lope Blanch, El Habla de la Ciudad
107).
'There are several things [that one understands...]'
Existential-Existential:
(14) Anteriormente no habian motores [qui hay ahora...] (Lope Blanch, El Habla
Popular 98).
'In the past there weren't motors [that exist nowadays...]'
Existential-Indirect Object:
(15) Y hay algunos otros [que...que les gusta la cosa arqueologica,...] (Lope Blanch, El
Habla de la Ciudad 441.
'And there are some other people [that...that like the archeological aspect,...]'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
209
Predicate Nominal-Circumstantial Complement:
(16) ...es la epoca [en que ya paso la floracion,...] (Lope Blanch, El Habla de la Ciudad
112).
'...in the season [in which the flowering has already passed,...]'
Circumstantial Complement-Existential:
(17 ) ...va uno a tomar refrescos y nieve al hotel Maris, alii, a la placita [que hay alii.]
(Lope Blanch, El Habla de la Ciudad 49).
'...one goes to have some refreshments and ice cream to the Maris hotel, there, to
the little square [that exist there.]'
Object of Preposition-Indirect Object:
(18) ...hay que ir con una persona.... [que Ie guste la historia del a rte,...] (Lope Blanch,
El Habla de la Ciudad 50).
'...you have to go with a person.-.Jto whom all those things are pleasing,...]'
Circumstantial Complement-Object of Preposition:
(19) ...mientras reciben su formation espiritual, sacerdotal en general, en el colegio [de
que se trata]... (Lope Blanch, El Habla de la Ciudad 56).
'...while they get their spiritual learning, pastoral in general, in the school [of their
selection]...'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 1 0
Indirect Object-Subject:
(20) ...como si a un...a alguien [que va a hacer una obra], se le entregan las
herramientas para hacerla. (Lope Blanch, El Habla de la Ciudad 57)
'...as if the necessary tools were given to a...to someone [who is going to make a
masterpiece]'
Predicate Nominal-Indirect Object:
(21) ...a este...por aca, que es una parte [que le dicen Parres] (Lope Blanch, El Habla
Popular 435).
'...to this...around here, which is a part [that they call Parres]'
S ubj ect-Existential:
(22) Una idea [que hay en Mexico,]...es el camino a Tepozteco,... (Lope Blanch, El
Habla de la Ciudad 195).
'An idea [that exists in Mexico,]...is the way to Tepozteco,...'
Object of Preposition-Circumstantial Complement:
(23) ...son gentes completamente distinto a lo que tu piensas o al medio [que tu tienes
que desenvolverte,...] (Lope Blanch, El Habla de la Ciudad 266).
'...they are people that are completely different from what you think or from the
environment [in which you have to function,...]'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 1 1
Object of Com parison-Subject3 1 :
(24) ...no me dejaron hacer como...asi, como otros. [que han cogido muchos terrenos.]
(Lope Blanch, El Habla Popular 193).
'...I was not allowed to do as...like that, as others, [who have taken several
terrains.]'
Indirect Object-Direct Object:
(25) Las compro a los comerciantes ahi [que conoce el.] (Lope Blanch, El Habla de la
Ciudad 186).
The traders [that he knows], he paid them for these things.'
Direct Object-Indirect Object:
(26) ...me dio la obra Landru opereta. [que Salvador Elizondo le puso la musica...]
(Lope Blanch, El Habla de la Ciudad 329).
'...she gave me the book Landru opereta. [to which Salvador Elizondo composed
gave the music...]'
Circumstantial Complement-Indirect Object:
(27) Hay un senor que vive alia, por el Penon Vieio. [que le nombran Aviacion.] (Lope
Blanch, El Habla Popular 441).
'There a man who lives there, by the Penon Vieio. [that people call Aviacion.]'
3 lWhen the speaker establishes a comparison of objects, persons and events by
using 'como/like' followed by a noun phrase, this noun phrase was considered Object of
Comparison.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2 1 2
Subject-Indirect Object:
(28) ...el gato. [a quien mi sobrina le tenia tambien un gran carino,] no aparecia... (Lope
Blanch, El Habla de la Ciudad 127).
'.-.the cat, [whom my niece also had a great affection for,] would not show up...'
Subject-Possessive:
(29) ...Por alia vive una senora [que hace muchos anos, a su papa lo metieron a la
carcel,...] (Lope Blanch, El Habla Popular 188).
'A lady [whose father was taken to jail several years ago,...] lives over there.'
Indirect Object-Indirect Object:
(30) ...ahi, a los eiidatarios [que nos entregaron aqui...] este...nos entregaron ... (Lope
Blanch, El Habla Popular 191).
'...there, they gave it to us, to the settlers [whom they gave here...]'
Object of Comparison-Existential:
(31) ...aqui no se velan casas como las [que hay orita.] (Lope Blanch, El Habla
Popular 274).
'...houses like the ones [that exist here] were not seen here.'
Predicate Nominal-Possessive:
(32) ...porque el es una persona [que toda su...sus...bueno, todos sus amigos de el,
pues...les gusta robar, les gusta fumar...] (Lope Blanch, El Habla Popular 75).
'...because he is a person [whose friends like to steal, they like to smoke...]'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
213
Object of Com parison-Direct O bject:
(33) ...como el trabaio [que tengo yo]... (Lope Blanch, El Espanol 281).
'...like the job [that I have here]...'
Object of Comparison-Circumstantial Complement:
(34) No es tanto como...como San Francisco, [que hay jtanta gente, tanta gente!] (Lope
Blanch, El Espanol 299).
'It is not like...like San Francisco, [in which there are so many people!]'
Direct Object-Possessive:
(35) Porque yo orita tengo una amiga [que su nino,... ] (Lope Blanch, El Espanol 289).
'Because now I have a friend [whose child,... ]'
Indirect Object-Object of Preposition
(36) Y en el [que tamos trabajando orita] le dicen el Peacekeeper,... (Lope Blanch, El
Espanol 273).
'And the one [that we are currently working on] they call the Peacekeeper,...'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
214
APPENDIX 2
The following types of registers are analyzed in this study:
1. One-to-one interview (directed dialogue).
2. One-to-one interview (free conversation).
3. One-to-one interview (free conversation, a third person participates).
4. One-to-one interview (secret recording).
4. One-to-one interview (directed dialogue with the presence of a third person who talks
occasionally).
5. Group interview (directed dialogue, interviewer participates).
6. Group interview (free conversation, interviewer participates).
7. Group interview (free conversation, interviewer does not participate).
8. Group interview (secret recording, interviewer participates).
9. Monologue.
10. Conferences/University Lectures.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
APPENDIX 3
215
The social variables in this study are the following:
1. Socioeconomic level
1.1. Middle class
1.2. Lower class
2. Level of education
2.1. Complete primary education
2.2. Incomplete primary education
2.3. Complete secondary education
2.4. Incomplete secondary education
2.5. Technical school
2.6. College education
2.7. Level of education not specified.
2.8. No education
3. Bilingual/Monolingual
3.1. Bilingual
3.2. Monolingual
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
APPENDIX 4
HUMANNESS OF ELEMENT THAT PROVIDES THE GROUNDING BY TYPE OF GROUNDING
_______________________ (NONHUMAN ANTECEDENTS)_____________________________
Humanness of element that
provides the grounding
TYPE OF GROUNDING TOTAL
Anchoring Main-Clause
Grounding
N %
N % N %
Element that provides the
grounding is Human
313 95% 441 90% 754 92%
Element that provides grounding is
NonHuman
18 5% 47 10% 65 8%
TOTAL 331 100% 488 100% 819 100%
p<,029
216
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
APPENDIX 4 (continued)
HUMANNESS OF ELEMENT THAT PROVIDES THE GROUNDING BY TYPE OF GROUNDING
__________________________ (HUMAN ANTECEDENTS)________________________________
Humanness o f element that
provides the grounding
TYPE OF GROUNDING TOTAL
Anchoring Main-Clause
Grounding
N %
N % N %
Element that provides the
grounding is Human
63 83% 177 89% 240 87%
Element that provides grounding is
NonHuman
13 17% 22 11% 35 13%
TOTAL 76 100% 199 100% 275 100%
p<.178
217

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

An anaphoric approach to clitic position in Spanish

PDF

Acquisition of Spanish verb morphology by bilingual children. A longitudinal study between the ages of 2;9 and 3;3

Register and style variation in speakers of Spanish as a heritage and as a second language

PDF

Discourse functional units: A re-examination of discourse markers with particular reference to Spanish

PDF

Conditions on null objects in Basque Spanish and their relation to leismo and clitic doubling

PDF

Adding and subtracting alternation: resumption and prepositional phrase chopping in Spanish relative clauses

PDF

The present perfect in Spanish: a study on semantic variation

PDF

The locative alternation: unaccusative constructions and subject position

PDF

Asymmetry of scope taking in wh -questions

PDF

Factors influencing the interpretation of novel words as adjectives in 4-year-old Spanish -speaking children

PDF

Asymetrical discourse in a computer-mediated environment

PDF

A discourse analysis of teacher-student classroom interactions

PDF

A Sociolinguistic Study Of Selected Vowel Changes In Los Angeles English

PDF

Gossip, letters, phones: The scandal of female networks in film and literature

PDF

Empires of the habanera: Cuba in the cultural imaginary of Catalonia

PDF

Prosody in contact: Spanish in Los Angeles

PDF

Child As Apprentice-Narrator: Socializingvoice, Face, Identity, And Self-Esteem Amid The Narrative Politics Of Family Dinner

PDF

Grammaticalization and the development of functional categories in Chinese

PDF

From the ruins of empire: Afrohispanic writings of Equatorial Guinea, the Caribbean and Spain, 1980--2000

PDF

The Spanish continuum in Peruvian bilingual speakers: A study of verbal clitics

Asset Metadata

Creator Gervasi, Kareen Liliana (author)

Core Title A variationist study of relative clauses in Spanish

School Graduate School

Degree Doctor of Philosophy

Degree Program Spanish

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag language, linguistics,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Silva-Corvalan, Carmen (committee chair), Finegan, Edward (committee member), Saltarelli, Mario (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-61003

Unique identifier UC11338976

Identifier 3018002.pdf (filename),usctheses-c16-61003 (legacy record id)

Legacy Identifier 3018002.pdf

Dmrecord 61003

Document Type Dissertation

Rights Gervasi, Kareen Liliana

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA