Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Kinship Terms

Download as pdf or txt
Download as pdf or txt
You are on page 1of 20

Ascending kinship terminology

in Middle Rocky Mountain English

Lamont D. Antieau

This study uses the tools of corpus linguistics to investigate ascending kin-
ship terminology in the Linguistic Atlas of the Middle Rockies, a collection of
interviews gathered in Colorado, Utah, and Wyoming as part of a dialectologi-
cal survey of the American West. Relying in part on the framework of Dahl
and Koptjevskaja-Tamm (2001), particularly with respect to their notion of a
parental kin prototype, the study examines lexical and grammatical variation in
the use of terms for parents and grandparents in different interviewing contexts
in an effort to identify patterns in these distributions. The study finds important
quantitative differences in the distribution of mother and father, as well as dif-
ferences in the grammatical behavior of these and other kinship variants. While
these results provide some support for a parental kin prototype, they also suggest
the benefits that survey data collected within a variationist framework offer such
a prototype, both with respect to the counterexamples to broad generalizations
that such datasets inevitably include as well as the variable patterns that often
emerge from such data that might go unobserved using formal methods.

Keywords: American English, language variation, sociolinguistics, dialectology,


corpus linguistics, kinship terminology, linguistic atlases, language typology

1. Introduction

Kinship terms have an important function in language in that they enable speakers
to mark their relationships with other members of their families explicitly via the
lexicon. At the same time, they constitute evidence of the social structure of fam-
ily organizations, and differences in their use can reveal varying levels of intimacy
among relations. Because they are used by speakers to refer to significant people
in their lives, kinship terms often constitute a large part of our daily discourse, and
as such, there has been an abundance of research on kinship terminology in the
fields of anthropology, linguistics, and sociology.

English World-Wide 33:2 (2012), 185–204. doi 10.1075/eww.33.2.04ant


issn 0172–8865 / e-issn 1569–9730 © John Benjamins Publishing Company
186 Lamont D. Antieau

Despite their importance in conversation and the wealth of research on


them, several facets of kinship terminology have gone unexplored. Dahl and
Koptjevskaja-Tamm (2001: 201; hereafter D&K), for instance, point out that the
grammatical properties of kinship terminology have not received the same sys-
tematic treatment given to those same issues in other lexical domains. With re-
gards to dialectal variation, while kinship terms have always been explicit targets
of the worksheets of the Linguistic Atlas of the United States and Canada, the dis-
cussion of this data has often focused on phonetic variation in the targets father
and daughter (e.g. Kurath and McDavid 1961: 112, 161) rather than on their lexical
distribution. Furthermore, much of the earlier Atlas descriptions focused on elici-
tations, in part because of limitations imposed by earlier methods of data collec-
tion, and, as such, ignored important aspects of their use. As Allen (1975: 338) not-
ed in his discussion of kinship terms in the Linguistic Atlas of the Upper Midwest:
Rarely discernible in the limited data are the well-known influences of the emo-
tional timbre of a situation — playful, matter-of-fact, affectionate, angry, etc.
— upon the choice of the term of address. Within the speech of one person the
shift from daddy to dad to father can readily occur as the mood of a conversation
changes, but such a change in register calls for a different kind of survey.

In an effort to better understand this complex lexical domain, then, the current study
relies on both elicited and conversational data, as well as metadata, to investigate the
lexical and grammatical properties of kinship terminology in the Middle Rocky
Mountains, a region that has received little coverage in the variationist literature.
To do so, the methods of dialectology were integrated with the tools of corpus
linguistics (see Anderwald and Szmrecsanyi 2009; Wagner and Anderwald 2007)
to analyze data collected for the Linguistic Atlas of the Middle Rockies (LAMR). The
study begins with a description of the framework created for studying the gram-
mar of kinship terms by D&K (2001) and then a summary of the LAMR project,
before presenting the kinship terms used by speakers in the LAMR corpus and the
grammatical distribution of the major variants. Although one of the primary mo-
tivations for doing so is simply to describe kinship terminology in a little-known
variety of American English, the work has two additional goals: 1) to compare the
language of kinship in Middle Rocky Mountain English to that of other varieties
of English, and 2) to show how work in dialectology and linguistic typology can
complement each other (in the spirit of, for example, Kortmann 2004), particu-
larly in the analysis and description of lexical domains, as represented in this study
by ascending kinship terms.
Ascending kinship terminology in Middle Rocky Mountain English 187

2. Background and terminology

In an effort to provide a foundation for studying variation in kinship terms, this


study adopts the framework and terminology of D&K (2001) in their work on
the grammar of kinship terms from a typological perspective. D&K (2001: 201–2)
begin by establishing two key components of kinship terms: 1) a referent denoted
by the kinship term; and 2) an anchor explicitly or implicitly linked to the referent
in the discourse. Thus, for the utterance “My mother couldn’t be here today”, the
referent is the person in the field of discourse to whom the term mother applies,
and the anchor is the speaker of the utterance, a role that is made explicit via the
possessive pronoun my. If the explicit link my were omitted, the referent and the
anchor would remain the same, but the relationship would be considered implicit.
Kinship terms are horizontal when both the referent and the anchor belong to the
same generation, descending when the generation of the anchor precedes that of
the referent, and ascending when the generation of the referent precedes that of
the anchor. Parental kinship terminology consists of ascending kinship terms that
refer only to the generation immediately preceding the anchor.
Another important component of D&K’s framework is the recognition
that kinship terms are put to use in a variety of ways and in different contexts
(2001: 202). A use of a kinship term is said to be egocentric when one or more par-
ticipants of the speech act are anchors (Dahl 1997), and it is in-family, as opposed
to out-of-family, when the other participants of the conversation are in the same
family as the anchor of the kinship term. D&K also make a distinction between
vocative, referential, and predicative uses of kinship terms, and they point out that
the focus in their work is on vocative and referential rather than predicative use
due to the nature of the Swedish corpus on which their analysis is based.
Although the current paper adopts much of the terminology of D&K and
shares many of the same goals, there are some differences as well, stemming in
great part from differences in the data that were used in each study. Contributors
to the Swedish corpus used in D&K’s study were asked to record a half-hour con-
versation with someone they knew well; as a result, some of the conversations
included members of the same family and consequently kinship terms emerged
as participants used them to address the other participants in the conversation
or to discuss something the participants had in common: family. On the other
hand, the Atlas interviews analyzed in the current study were conducted one-on-
one between a fieldworker and informant who were typically strangers outside the
context of the interview. Because Atlas informants tend to be elderly themselves,
it is rare that their ascending kin are alive, let alone present during the interview;
thus, authentic vocative uses between speaker and parental referents are rarely
found in the corpus, and informants are generally restricted to talking about their
188 Lamont D. Antieau

parents and grandparents, rather than talking to them. However, Atlas interviews
address a dimension that is not accounted for systematically in the Swedish corpus
but relates to vocative usage: terms of address for parents and grandparents are ex-
plicit targets in the Atlas worksheets, relying on prompts of the form, for example,
“What did you call your grandmother when you were growing up?”.
While access to elicited data provides the study a broader view than it would
have without such data, in other ways the current study takes a more restricted
approach to kinship terms than D&K and others do, not only by focusing on a
geographically and socially restricted variety of American English, but in that the
speech acts being investigated are egocentric (the informant is the anchor) and
out-of-family (the field worker is a stranger). Furthermore, the focus is on ascend-
ing kinship terms rather than descending or horizontal because the Atlas work-
sheets that guided the interviews explicitly targeted mother and father (worksheet
items A1 and A2, respectively), as well as grandmother and grandfather (A4 and
A5), while only explicitly targeting daughter among descending informants (A22).
Additionally, the parental prototype proposed by D&K (2001: 216) and their dis-
cussion of doublets, with its focus on parental terms, are two issues also of concern
to the current study that guided it in narrowing its focus to ascending terms.

3. Methods

The data for this study were extracted from interviews conducted as part of the
LAMR project, a subcomponent of the larger Linguistic Atlas of the Western States
(LAWS). The general methods of the LAWS project have been described elsewhere,
both in terms of exhaustive detail covering all aspects of the project (e.g. Pederson
and Madsen 1989; Pederson 1990, 1996b; Hamilton-Brehm 2003a; Antieau 2006)
and in specific aspects of the methodology (e.g. Pederson 1996a; Barry and
Antieau 2001). However, it is worthwhile to note a few general characteristics of
the work here. As a regional survey of language use, the project seeks broad cover-
age via field workers who identify rural communities throughout a region as being
worthy of interest; locate informants who were born or lived most of their lives in
these communities; and engage in a three-hour, audio-recorded interview with
informants to collect data on language use and culture in their respective commu-
nities. As in earlier Atlas surveys, informants tend to be elderly, at least for the first
round of interviewing, and have farming and/or ranching experience.
The most significant difference between LAWS methods and earlier Atlas
methods is that interviews are audio-recorded, thereby capturing all audio aspects
of the interview, the speech of which is then fully transcribed in standard orthog-
raphy, allowing for the results of field work to be investigated using the tools of
Ascending kinship terminology in Middle Rocky Mountain English 189

W E Wyoming

Utah

Colorado

Map 1. LAMR informants by community location

corpus linguistics. At the time of analysis for this paper, 66 of the 70 interviews
collected in the middle Rocky Mountain states had been transcribed in their en-
tirety and were therefore used in the analysis. A map showing the locations of the
informants is presented above.
Of the 66 informants, 15 lived in Wyoming, 17 in Utah and 34 in Colorado.
56 of these interviews are considered primary in that they are the most complete
field records that LAMR has for those communities; ten additional interviews
were conducted in some of these communities as well but are considered auxiliary
in that they are incomplete and/or had audio issues. Interviews were conducted
during two timeframes: 48 were conducted from 1988 to 1991 and 18 from 2001
to 2004, with the latter all being conducted in Colorado. The ages of informants
ranged from 92 to 30 with a mean age of 75; informants’ years of birth ranged from
1898 to 1957. Finally, informants’ education levels ranged from 8th grade to post-
graduate college work; in all, 20 informants reported attending college or business
school, 23 graduated high school but did not continue their education, 20 did not
graduate from high school, and the level of education that three informants at-
tained is unrecoverable from the interviews.
Data analysis relied on Kwickwic1, a concordance tool that accepts linguis-
tic features as queries, retrieves those features from document collections and

1. Developed by Clayton Darwin and available at <http://www.kwickwic.com>.


190 Lamont D. Antieau

then presents passages containing those linguistic features as keywords in context


(KWICs). Initial queries comprised kinship terms that were known or assumed to
appear in the corpus such as papa, dad, mommy, and granddad. Since the number
of headwords investigated as part of this study was relatively small, all the extract-
ed KWICs were read to ascertain that the variants collected were egocentric, that
is, that they were used to refer to the parents and grandparents of that informant
and not, for instance, to those of their friends or as parts of metaphors or idioms
such as daddy cow and mama’s boy. An iterative approach was taken to the data
in that, as additional kinship terms emerged from reading chunks of individual
interviews, they were rerun on the entire dataset.
An initial survey of the corpus revealed that kinship terms appear quite fre-
quently in the LAMR corpus, reflecting the historical nature of the interview for-
mat and its focus not only on the family but on topics like the home, the com-
munity, childhood activities, etc. In the field record from Rock River, Wyoming,
for instance, the term mother appears 73 times over the course of the interview.
Many of these uses provide redundant information for the purposes of this study,
and thus a type-token method that counted all occurrences of kinship terms was
eschewed in favor of an approach that established the presence or absence of spe-
cific lexical variants across a range of grammatical constructions appearing in in-
dividual interviews.
In those cases in which an ascending kinship term was validated as egocentric,
the following characteristics of the use of the term in each interview were noted: 1)
whether the feature in question was elicited by the fieldworker or occurred natu-
rally in conversation; 2) what the morphosyntactic properties of the term were,
e.g. whether the term was in genitive case, appeared after a first-person possessive
pronoun, or appeared in a coordinate construction with another kinship term; and
3) whether the informant provided any relevant metadata concerning the use of
the term.

4. Results

The results are presented here in three sections. The first section comprises data
on the range of lexical variants pertaining to parents and grandparents that were
either elicited by fieldworkers or emerged during conversational portions of the
interview,2 and the second section details the grammatical constructions associ-
ated with primary variants culled from the lexical analysis.

2. The ascending kinship headwords aunt and uncle are not analyzed in Section 4.1 because
neither was prone to lexical variation; however, they will be discussed in Section 4.2 with respect
Ascending kinship terminology in Middle Rocky Mountain English 191

4.1 Lexical variation

With the exception of an auxiliary interview conducted in Brush, Colorado, which


covered only the last half of the worksheets, ascending kinship terms appeared in
all of the LAMR interviews analyzed here. The following tables show patterns of
distribution in ascending kinship terminology as elicited variants (Table 1) and
conversational variants (Table 2). Each variant is followed by the number of speak-
ers in the corpus who used it in parentheses, with more than one term possible for
each speaker in both contexts.

Table 1. Elicited variants


Headword Variants
dad (29), daddy (10), papa (6), pop (2), J.V. (1), Jimmy (1), king of
Father the hill (1), old gobbler (1)
mom (26), mother (21), mama (8), ma (5),
Mother mommy (1), the boss (1)
grandpa (14), granddad (4), grandfather (3),
Grandfather jidow* (1), pa (1), papa (1), pop (1)
grandma (20), grandmother (3), gram (1),
Grandmother grammy (1), ma (1), sipaw (1)
* Jidow ‘grandfather’ and sipaw ‘grandmother’ were elicited from a native speaker of English who used the
terms as a child to address her grandparents in their native language (Lebanese).

Table 2. Conversational variants


Headword Variants
Father dad (60), father (53), daddy (8), dads* (1), Jimmy (1), papa (1), pop (1)
Mother mother (62), mom (26), mama (4), mommy (1)
Grandfather grandfather (45), grandpa (24), granddad (14)
Grandmother grandmother (41), grandma (21)
(great) great-grandfather (11),
Great- grandfather (great) great-granddad (2)
Great-grandmother great-grandmother (5)
* Dads was used multiple times by a single informant not as a plural form but with respect to a singular
referent, perhaps by analogy with such variants as pops or gramps.

To summarize, these tables show that the variants obtained for each of the major
target items, viz. father, mother, grandfather, and grandmother, are distributed ac-
cording to a power law with two or three variants used by the majority of speakers
for each headword, followed by several low-frequency items. Thus, the distribu-
tion of ascending kinship terms in the LAMR corpus follows the same pattern

to grammatical variation in their usage.


192 Lamont D. Antieau

described in other studies of lexical variation in language use (e.g. Zipf 1949;
Kretzschmar and Tamasi 2003; Kretzschmar 2009, 2010; Burkette 2001, 2009).
Tables 1 and 2 show that this pattern of variation is the norm for both elicited
and conversational contexts, with the widest range of variation occurring in the
elicited context, presumably because such a narrow context allows for the use of
variants that would otherwise be vague and thus require explanation or be prone
to misinterpretation in freer types of discourse. Note, too, that several variants that
can refer to either father or grandfather, viz. pa, papa, and pop, are nearly exclusive
to the elicited context, with only pop being used in conversation and then only by
a single speaker.
In Table 3, pairs of parental variants appearing in coordinate constructions
are presented, followed by grandparental variants in Table 4, with the number of
speakers who used these phrases in conversation and elicitation. For most of the
variants in Table 3, the parental terms could appear in either order, that is, mother
and dad = dad and mother; these variants were collapsed into a single type repre-
senting both orders in this table.

Table 3. Ranking of coordinate parent constructions


Rank Variant Elicited* Conversation Total speakers**
1 mother and father 3 17 19
2 mother and dad 6 16 18
3 mom and dad 7 9 14
4 mother and daddy 4 — 4
5 mama and papa 3 — 3
6 mama and daddy 2 — 2
7 mother and Jimmy 1 — 1
* The prompt for this was a question along the lines of “Did you have another term for your parents?” The
intended targets for this prompt were synonyms for parents as a unit such as parents or folks; however, as
shown in this data, some informants provided coordinate constructions comprising individual parental
terms. Some fieldworkers, too, attempted to elicit both variants at the same time with a prompt like, “What
did you call your mother and father when you were growing up?”
** These totals reflect that an individual speaker could produce the same variant in both conversational
and elicited contexts and could use more than one kind of coordinate construction for parents.

As shown, the highest ranking variant is mother and father — a construction com-
prising parental variants mirroring each other in length, phonological structure
and formality. In a similar fashion, the variants ranking third, fifth, and sixth —
mom and dad, mama and papa, and mama and daddy, respectively — also mirror
each other in terms of their length, phonological structure (none more so than
the reduplications mama and papa), and formality (or lack thereof). The variants
ranking second, fourth, and seventh, however, adhere to an alternate pattern that
will be addressed in Section 5.
Ascending kinship terminology in Middle Rocky Mountain English 193

In Table 4, individual kinship terms could appear in the coordinate construc-


tion in either order, with the ordering here reflecting that, as a rule, most coordi-
nate grandparent constructions in the corpus are introduced by variants referring
to male grandparents.

Table 4. Ranking of coordinate grandparent constructions


Rank Variant Elicited Conversation Total speakers
1 grandpa and grandma 6 3 9
2 grandfather and grandmother — 5 5
3 granddad and grammy 1 — 1
4 granddad and grandmother — 1 1

As in the results of Table 3, the top-two constructions, viz. grandpa and grandma
and grandfather and grandmother, comprise terms that mirror each other in sev-
eral ways. The lower-ranking variants, however, pattern differently, which will also
be taken up in the discussion in Section 5.
In the next section, the focus will be on grammatical variation in those vari-
ants used by multiple speakers and will particularly focus on the single-term vari-
ants presented in Tables 1 and 2.

4.2 Grammatical variation

Grammatical analysis of ascending kinship terms began with the results of the
lexical analysis, in that only those variants of headwords that were used by mul-
tiple informants, as presented earlier, were used in the investigation into how such
terms behave grammatically. As in the analysis of lexical variants, a type-token
approach taking into account every grammatical construction that kinship terms
appear in the corpus was eschewed in favor of a categorical approach that explored
whether or not variants appeared in several specific grammatical constructions.
To do so, the analysis focused on conversational data, based on the assumption
that such data provides a more natural, and thus better, context for examining
grammatical variation than elicitations do. Following D&K, the present study fo-
cused on kinship terms appearing in two different cases — nominative and geni-
tive3 — and for each case, whether the kinship term: A) was a bare noun, e.g. “But
Mother was certainly a good nurse”; B) was preceded by a possessive pronoun, e.g.
“My father was a genius at mechanics”; C) was a bare noun followed by a name,

3. The investigation of genitive case in D&K’s study and the current one, however, is different
in that D&K were interested in kinship terms in the role of possessed, as in e.g. “John’s mother
…”; however, since the data for this study is limited to egocentric uses, I focus on kinship terms
in the role of possessor, as in e.g. “my father’s people …”.
194 Lamont D. Antieau

typically a last name for a grandparent, e.g. “And Grandfather Smith4 I know came
from England”, and a first name for an aunt or uncle, e.g. “Then Aunt Mary lived
right next to us”; or D) was preceded by a possessive pronoun and followed by a
name, e.g. “My Grandmother Jones never did use a separator”. Results for both the
nominative and genitive cases were lemmatized and tabulated together to produce
the table below.

Table 5. Grammatical variation in the conversational use of kinship terms


Grammatical Types
Headword Variants Total speakers
A B C D
father    4 49 — — 53
Father dad 35 56 — — 60
daddy    7    3 — — 8
mother 29 63 — — 63
Mother mom 20 15 — — 28
mama    3    1 — — 4
grandfather    7 43 3 3 45
Grandfather grandpa 13    6 8 3 23
granddad    5 10 3 1 11
grandmother    8 40 1 4 46
Grandmother
grandma 10    7 6 2 20
great-grandfather    1    8 — 2 10
Great-grandfather
great-granddad —    1 — 1 2
Great-grandmother great-grandmother —    5 — — 5
Aunt aunt — 14 5 1 18
Uncle uncle    1 21 5 6 33
Totals 140 342 31 23

Several patterns are apparent in the table. First, while many LAMR informants use
kinship terms in the bare form (Type A), more than double that number use such
terms preceded by the possessive pronouns my or our (Type B), as shown by the
total ratios of Type A and Type B constructions. No terms appear in only Type A
constructions without appearing in Type B constructions, although the opposite
case is realized in the distribution of the terms great-granddad, great-grandmother,
and aunt. However, there are some terms in which the number of speakers using a
variant in Type A constructions exceeds the number of speakers using that term in
Type B constructions, as in mom, mama, grandpa, and grandma. The latter state-
ment underscores the finding that speakers use the bare form for more formal
terms like father, mother, grandfather, grandmother, and great-grandfather far less

4. Pseudonyms are used in place of real names in the LAMR examples presented in this paper.
Ascending kinship terminology in Middle Rocky Mountain English 195

than they do for variants like dad, daddy, mom, grandpa, granddad, and grandma,
which suggests the more proper-name-like quality of the latter variants.
Terms at the parental level differ from terms at greater ascension (and thus,
greater distance from the anchor) in that the names of parents are never com-
bined with a kinship term, for example, a variant like Mother Barb is unattested;
thus, there are no instances of parental variants falling into Types C and D, while
several speakers used the kinship term + name template for either grandparent,
aunt, and uncle. This is presumably because while speakers have a single set of
biological parents, they have two sets of biological grandparents, and potentially
several aunts and uncles, thus requiring names to distinguish between referents
when the same kinship terms are to be used for them in discourse. One infor-
mant (in Meeker, Colorado) mentioned a distinction in how she referred to her
grandparents in this regard: for her maternal grandmother, she simply used the
term grandma; for her paternal grandmother, she used the term Grandma Davis.
However, there was not enough data in this collection to investigate whether there
were tendencies concerning the use of bare forms for one set of grandparents and
not the other set.
As the table shows, when LAMR informants use a kinship term in conjunction
with a name, as in Aunt Mary or Grandpa Smith, they are generally more likely to
use it as a bare form (Type C) than to introduce it with a possessive pronoun (Type
D); however, these results are mixed. The table also shows that more informants
use a bare kin term followed by a last name when the kin term is a less formal one
such as grandpa, granddad, or grandma than with the more formal terms grandfa-
ther or grandmother, a point I will return to in the next section.

5. Discussion

In an overview of lexical typology, Koptjevskaja-Tamm (2008: 20) says that “[c]las-


sifications become particularly interesting when they show various asymmetries
— statistical preferences for certain combinations of parameters and the absence
of attested though logically possible types”. In the current study, such asymmetries
are perhaps most apparent with respect to differences in the lexical distribution of
the variants of the headword mother and father, as shown in Tables 1 and 2 above.
Although there is clearly competition among formal and informal variants for
both the headwords father and mother, the contexts in which these variants com-
pete are different. Mother and mom compete in both contexts; however, while the
competition is relatively close in the elicitation context with mom ranking slightly
higher than mother, mother is used by nearly all informants and mom by fewer
than half that number in the conversational context. On the other hand, father and
196 Lamont D. Antieau

dad only compete in the conversational context, and despite its strong showing
in conversation, father never appears as an elicited form with several informants
even going on-record to deny having used the variant father as an address term, as
illustrated in the following interview:
(1) Excerpt from Alva, Wyoming:5
P: Now, you mentioned your mother and your father. Did you call them by
any other names?
R: We called them dad and mother.
P: Okay. Did you ever call your dad anything else? Or your mother, did you
ever call her anything else? Or was it always mother?
R: Well, mom, I think, more when we were little and ….
P: And you called your dad father occasionally or always dad?
R: It was always dad, I think, wasn’t it?
S: Yes, as far as I know.
R: I can’t remember. I don’t think I ever called him father.
S: I don’t think so. I don’t think I ever did either.

Further support was offered in other interviews, such as the following.


(2) Excerpt from Farmington, Utah:
P: That’s sad. What about your parents? Just call them mom and dad?
R: Pretty well. We called them mother and dad you know.
P: Okay. Was there another more formal way of approaching your father,
your dad?
R: No. Not too much. He was pretty common ordinary guy.
P: (laughs)
R: So we just, dad and mom was usually what we called them.
P: But is there a more formal word for dad?
R: Not really. You mean father?
P: Yeah.
R: Or something like that?
P: Yeah. That’s what I’m (interrupted)
R: Oh very seldom. I don’t think we ever called him the father.

Thus, in terms of elicitation, the variants father and mother produced very differ-
ent results among the LAMR informants, and some informants were conscious of
this difference. However, as self-reporting has produced results of questionable
value on occasion (cf. Chambers 1995; however, see Bailey, Wikle and Tillery 1997
for another perspective), we turn to other resources for support on this matter.

5. In interview excerpts, P=prompt of fieldworker, R=response of primary informant, and


S=response of spouse or other secondary informant present during the interview.
Ascending kinship terminology in Middle Rocky Mountain English 197

First, in addition to providing negative evidence for the use of father, both of
the excerpts above also briefly address an issue concerning the use of variants of
mother and father and their use in coordinate constructions. As shown in Table 3
above, mother and father is the highest-ranking coordinate construction com-
prising parental variants in the corpus, suggesting that kinship terms that mirror
each other might be more likely to appear together in coordinate constructions,
a claim that is given additional support by the third-ranking variant, mom and
dad. However, the constructions mother and dad, mother and daddy, and mother
and Jimmy, ranked second, fourth, and seventh, respectively, create problems for
such a generalization, and thus, an analysis of the entire coordinate construction
set instead suggests something along the lines of the following implicational gen-
eralization: if the parental terms in a coordinate construction do not mirror each
other in terms of formality, then the least formal will be the variant for father with
an alternative interpretation being that terms for mother in such constructions are
more invariant than terms for father.
The grammatical behavior of mother and father, as shown in Table 5, also
supports the idea that the two items are treated very differently by informants.
Specifically, father is used by only four informants as a bare form, while mother is
used as a bare form by 29 informants and overshadows the performance of mom
(20 speakers) as a bare form. Dad is used as a bare form in conversation by more
informants (35) than any other variant of the headword father, and in this respect,
patterns similarly to mother, and the two variants often appear in discourse to-
gether in the bare form, as shown in the example below.
(3) Excerpt from Rock River, Wyoming:
It was really something. Well, anyhow, Dad sent for Mother to come out in
November. And she came out, and he met her in Laramie. (false start) There
was a passenger train, you know, through Rock River then, at that time. And
he met her there, and they were married, and they came to Rock River. Well,
Dad was quite a well-known cowboy here, and he was a very jolly man. And
so he met her …

Thus, elicited data, coordinate constructions, quantitative differences in bare


forms, testimony from informants and anecdotal evidence suggest that the vari-
ants mother and father are evaluated differently by LAMR informants. More spe-
cifically, the LAMR data are in agreement with Atwood’s speculation concerning
the use of kinship terms in Texas: “It is somewhat doubtful that Father is often
used in real family situations” (Atwood 1962: 66). Rather, dad is generally used to
fill the role as variant for male parent in such situations in the LAMR data. While
the psychological and social reasons for the lack of use of the variant father in these
contexts is beyond the scope of this study, the remarks made by some informants
198 Lamont D. Antieau

suggest the word to be too serious and formal for use as an address term for their
male parent in most contexts. Furthermore, some informants use Father as a vari-
ant of God in the corpus, and thus, this use of the term could limit its appearance
in other contexts or its application to anything other than the religious referent.
Finally, we might be interested in how variants for parents and grandparents
compare, that is, whether the distributional properties of parental terms might
percolate, as it were, to the level of grandparents. As the table shows, variants for
grandparents generally follow the pattern set by variants of parents in that in the
conversational context, the formal terms grandfather and grandmother are used by
the greatest number of informants. In the context of elicitation, again conforming to
the general pattern set by parent variants, the informal terms grandpa and grandma
rank highest. In the details, however, grandparent variants favor both sides of the
family. Unlike the performance of mother in the elicitation context, neither grand-
father nor grandmother can be said to compete truly with the informal variant;
however, unlike father, these variants do appear in the elicited context, if by only a
few speakers. That both sets of grandparent variants fall in-between the extremes
of father and mother is also reflected in their grammatical behavior: as shown in
Table 5, grandfather is used in its bare form by seven out of 46 informants (16%)
and grandmother by eight out of 46 (17%), as opposed to the weak performance of
father in this respect (4/53, or 8%) and the strong performance of mother (29/63,
or 46%). Additionally, evidence in the clustering of grandmother and grandpa in
chunks of discourse, as in (4), is not as strong as was the case for mother and dad,
with grandma and grandpa often co-occurring, as in (5) and (6) below:
(4) Excerpt from Rock River, Wyoming:
Well, on my father’s side, I think they came from Ireland or England in
there, and I have contacted some of the elder (redacted name), and they tell
me that they mostly were all butchers, you know. And when they came to
United States, why, they, most of them, went into the butchering business.
Now, what Grandpa did, I’m not sure (false start) because my dad only
knew him for, you know, never knew him really because he was only six
months old when he died. And then Grandmother, of course, was so busy
raising the children, I guess, that he never inquired. But I do know that he
knew Jefferson Davis. Old Jeffy used to ride a mule, and he’d stop by my
grandmother’s and would give my dad a ride on the mule. Yes, so …. But,
anyhow, Dad didn’t have much of an education at all. And his English was
rather poor, really, and it was very hard for the children, you know.
(5) Excerpt from Leadville, Colorado:
Well Grandma was a housewife all her life, Grandpa was a, he worked on
the railroad, he repaired (hesitation) coal street operator in Malta, and he
Ascending kinship terminology in Middle Rocky Mountain English 199

was, what was it, he checked the brakes and oiled the car trains. Oh, I guess
that he was post office. He was a postmaster in Malta. He was the second
postmaster down there and it was something like 1887.
(6) Excerpt from Jefferson, Colorado:
One of the teachers who taught over at the little Fremont School lived with
Grandma and Grandpa, or lived with Grandpa, and, oh, I guess Grandma
was there at that time, too. The second grandma. And she had appendicitis
and she was only sick three days and she died right there at the ranch and
many, many people died with appendicitis.

However, coordinate constructions offer little evidence of grandparent variants


exhibiting the same mix of formal and informal variants that parental terms do
in coordinate constructions, and there is nothing in the way of informant testi-
mony regarding choices of grandparent variants to provide evidence that one or
the other might be avoided in family situations.
A comparison of parental and grandparental data does, however, raise an issue
concerning Atwood’s (1962: 66) speculation on father not being used in family sit-
uations. Although, like Atwood’s informants, LAMR informants also seem to have
little use for the variant father in such contexts, the implication that father is the
odd-man out, as it were, is questionable based on this data; rather, mother could ac-
tually be characterized as patterning differently than the other formal terms inves-
tigated here. At any rate, the LAMR data provide clear differences in the lexical and
grammatical choices that speakers make when referring to each of their parents.
Naturally, before commencing to compare the lexical and grammatical distri-
bution of ascending kinship terms in Middle Rocky Mountain English to a kin-
ship prototype in the world’s languages, a comparison of the nature of such terms
in the LAMR corpus with similar data collected in other regions of the United
States is in order. As such, I examined three studies providing evidence of lexical
distribution of kinship terms elicited from speakers in other regions of the United
States: Hamilton-Brehm (2003b), Johnson (1996), and Allen (1975). Hamilton-
Brehm (2003b) presents findings illustrating the same high use of mother (50% of
her informants) and low use of father (7.5%) among her El Paso informants that
was reported for informants in this study. In her study of variation and change
in the lexicon of the southeastern United States, which relied on early LAMSAS
data from the 1930s and her own fieldwork in 1990, Johnson (1996: 253) shows
the same avoidance of father as an address term among informants in both the
1930s and 1990 as LAMR does; grandfather is also rarely used by Johnson’s 1990
informants, who instead preferred grandpa, despite grandfather being the most
common variant used by her speakers from the 1930s. Johnson’s data for the head-
word grandmother is also similar to the LAMR findings, with grandma being the
200 Lamont D. Antieau

most common term and grandmother the next most common in both the 1930s
and 1990 data. Finally, Allen (1975: 338–9) found a similar avoidance of father in
his data from the Upper Midwest, where informants preferred the variant dad;
mother is also the most common variant reported as a term of address by infor-
mants in the field. For grandparent variants, Allen found that “[a]s a term of ad-
dress in the [Upper Midwest] grandfather, though found in all five states among all
three inf. types, is not so general as is the less formal and often more affectionate
grandpa. Third in frequency is familiar granddad …” (Allen 1975: 340). He contin-
ues in Section 50.2:
Terms for one’s grandmother are much like those for a grandfather …. The more
dignified grandmother exists as the sole expression for some infs., a formal alter-
nate to a familiar term for others, and only a reference word, rather than a voca-
tive, for a few …. The widespread common term is grandma (Allen 1975: 340).

Such findings are very similar to those of the LAMR investigation, despite differ-
ences in time and space, as well as methodological differences between some of
the projects.
However, there are some apparent differences between the results of these
studies and LAMR as well. Hamilton-Brehm (2003b) reports daddy (47%) and dad
(40%) as the top-ranking variants for the headword father, which contrasts with
the ranking they hold in the LAMR data; Johnson (1996: 254) reports an over-
whelming preference for daddy as an address term in her 1990 data as well, while
papa, daddy, and pa compete for the top ranking in the 1930s data she used. These
studies taken together suggest an overall preference for the variant daddy in the
southern United States.6 Johnson (1996: 254) also reports preference for the term
mama for the headword mother in the 1930s and 1990, with mother appearing as
the second-highest ranking variant in the 1990 data. In the Upper Midwest, Allen
(1975: 338) reports papa being the second-ranking variant after dad for the head-
word father, although in the mail checklists he used, he found that daddy barely
edged out papa for the second spot. In the responses to address terms for mother,
Allen (1975: 339) finds that after mother, mama was the second highest-ranking
variant, although in his mail checklist mom was the clear favorite, followed by
ma, mama, and mother, which he attributes to a recent innovation. Interestingly,
Allen likens the pervasiveness of mother and mama as variants of mother to dad
and papa as variants of father, when he says, “[t]he general observation made for

6. However, based on evidence in their respective surveys, Allen (1975) and Hamilton-Brehm
(2003b) reach contrasting conclusions about the regional origins of daddy, with the former
speculating that it is a northern form, and the latter a southern form, although Allen does report
that “instances are too few to warrant a firm conclusion” (1975: 338). Such contrasting views
suggest the need for more data and further research on the topic.
Ascending kinship terminology in Middle Rocky Mountain English 201

father … applies as well to the terms of address to one’s mother” (1975: 339), ap-
parently ignoring the replacement of the variant father by dad in this paradigm.
Unfortunately, none of these three studies presented findings on grammatical vari-
ation in kinship terminology, as such data might provide additional insight into
similarities and differences between the use of kinship terms in the middle Rocky
Mountains and elsewhere in the United States.
Returning to the parental prototype proposed by D&K (2001: 216), the data
in the LAMR corpus generally adhere well to the model. First, a kinship term in
the corpus is more likely to be used as proper-name-like, that is, appear in its bare
form, if it has a unique referent in the family, e.g. mother or father rather than aunt
or uncle; thus, while variants of mother and father often appear in their bare forms,
there is only a single instance of uncle in the bare form for either the nominative
or genitive case, and the remaining instances of aunt and uncle are preceded by a
possessive pronoun (Type B), followed by a first name (Type C), or both (Type D).
Grandparent variants also surfaced in their bare form, however, typically after be-
ing introduced into the conversation with their surnames. Additionally, several
informants expressed knowing only one set of grandparents, thus rendering con-
fusion between the two sets of grandparents and requiring surnames to stem that
confusion less likely.
Second, the evidence presented in Table 5 supports the condition of the pro-
totype that kinship terms are more likely to be proper-name-like if their referent
is no more than a single generation from the anchor. Thus, over half the speakers
who use variants of mother and father use them in their bare form at some point
of the interview, in both nominative and genitive cases, while fewer than half the
speakers who use variants of grandfather and grandmother do so as bare forms,
also using a possessive pronoun (Type B), a last name (Type C) or both (Type D)
when using these terms in conversation. Furthermore, the performance of these
variants in conversation is supported by their appearance in elicitation contexts,
coordinate constructions, and in the metadata that occurs with regard to some
variants in some interviews.
There is also evidence in this study of the kind of variation that D&K express
as “kin doublets”, although, as shown in several tables, the headwords father,
mother, and grandfather present three major variants each, followed by the hapax
legomena typical for such sets of lexical items. More importantly, there is evidence
supporting D&K’s (2001: 218) finding that these synonyms differ with respect to
use, such that it is typical for one variant to be used for certain functions, e.g. voca-
tive uses, and not in others, e.g. referential uses, and that the grammatical behavior
of variants corresponds to these different functions, as they illustrate with the be-
havior of père and papa in French:
202 Lamont D. Antieau

(7) ?? Père est venu \ Mon père est venu \ Papa est venu.

Differences between Tables 1 and 2 above shed some light on the lexical distribu-
tions of LAMR variants in this regard, as several variants for ascending kin appear
only in the contexts of elicitation and never in conversation. The lack of symmetry
regarding the use of the variants father and mother as bare forms and in coordinate
constructions also offers evidence that not all kinship terms, not even ones with
phonetic and semantic similarities, are treated the same either in terms of their
lexical or their grammatical properties. It is important to note, however, that in
the LAMR collection such distinctions among competing variants are generally
quantitative rather than qualitative. That is, rather than variants never appearing
in certain contexts, as D&K claim to be the case for papa in French, illustrated
in (7) above, Rocky Mountain speakers may use competing variants in different
contexts, with even the same speaker sometimes alternating, as in the following
examples from the Durango, Colorado, informant:
(8) Father came from eastern Ireland \ My father came earlier than that \ My
dad came with him and they signed their name …

Despite the variability of such terms in LAMR, however, there are certainly strong
tendencies for some variants to surface in certain grammatical contexts and not
others (see, for example, the distribution of father, grandfather, and grandmother
as bare forms in Table 5), and of course there is the total absence of father as an
elicited form, which was addressed earlier.
As a potential addition to the parental prototype, and something that D&K al-
lude to in their discussion of papa in French, the LAMR data suggests a correlation
between the formality of a term and its use as a proper-like-name such that the
more formal a term, the less likely it is to be used as a proper-name-like variant.
This is illustrated most clearly in the weak performance of father as a bare form, an
elicited form, and in coordinate constructions with mother, as well as its character-
ization in the metadata of several interviews. While the performance of the variant
mother under these same conditions poses a challenge to this potential addition to
the prototype, variants for grandparents offer additional support to the idea that
informal terms are more likely to be used as proper-name-like in conversation.

6. Conclusion

This study used the tools of corpus linguistics to investigate the lexical and gram-
matical properties of ascending kinship terminology in a set of Atlas interviews
collected in the middle Rocky Mountain states. The results of this study support
Ascending kinship terminology in Middle Rocky Mountain English 203

several of the observations made by D&K (2001) toward a kinship term prototype,
although it suggests some qualifications to general statements are in order to ac-
count for some of the complex variation that dialectological research inevitably un-
earths. Just as importantly, the study also suggests ways in which survey results have
the potential to make positive contributions to more formal linguistic constructs.

References

Allen, Harold B. 1975. The Linguistic Atlas of the Upper Midwest. Vol. 1. Minneapolis: University
of Minnesota Press.
Anderwald, Lieselotte and Benedikt Szmrecsanyi. 2009. “Corpus linguistics and dialectology”.
In Anke Lüdeling and Merja Kytö, eds. Corpus Linguistics: An International Handbook.
Berlin, New York: Mouton de Gruyter, 1126–39.
Antieau, Lamont D. 2006. “A distributional analysis of rural Colorado English”. Ph.D. disserta-
tion, University of Georgia.
Atwood, E. Bagby. 1962. The Regional Vocabulary of Texas. Austin: University of Texas Press.
Barry, Betsy and Lamont D. Antieau. 2001. “The Linguistic Atlas Spoken Corpus: the next gen-
eration in Atlas research”. Paper presented to the American Dialect Society, Atlanta, GA, 10
November 2001.
Bailey, Guy, Tom Wikle and Jan Tillery. 1997. “The effects of methods on results in dialectology”.
English World-Wide 18: 35–63.
Burkette, Allison. 2001. “The story of chester drawers”. American Speech 76: 139–57.
———. 2009. “The lion, the witch, and the armoire: lexical variation in case furniture terms”.
American Speech 84: 315–39.
Chambers, J.K. 1995. Sociolinguistic Theory. Oxford: Blackwell.
Dahl, Östen. 1997. “Egocentricity in discourse and syntax”. <http://www.ling.su.se/staff/
oesten/egocentric> (defunct 21 Dec. 2011).
——— and Maria Koptjevskaja-Tamm. 2001. “Kinship in grammar”. In Irène Baron, Michael
Herslund and Finn Sorenson, eds. Dimensions of Possession. Amsterdam, Philadelphia:
Benjamins, 201–25.
Hamilton-Brehm, Anne Marie. 2003a. “A foundational study of El Paso English”. Ph.D. disserta-
tion, University of Georgia.
———. 2003b. “Substantial evidence of lexical variation in El Paso, Texas”. Paper presented to the
American Dialect Society, Atlanta, GA, 2 January 2003.
Johnson, Ellen. 1996. Lexical Variation and Change in the Southeastern United States, 1930–
1990. Tuscaloosa, AL: The University of Alabama Press.
Koptjevskaja-Tamm, Maria. 2008. “Approaching lexical typology”. In Martine Vanhove, ed.
From Polysemy to Semantic Change. Amsterdam, Philadelphia: John Benjamins, 3–52.
Kortmann, Bernd, ed. 2004. Dialectology Meets Typology: Dialect Grammar from a Cross-
linguistic Perspective. Berlin, New York: Mouton de Gruyter.
Kretzschmar, William A., Jr. 2009. The Linguistics of Speech. Cambridge: Cambridge University
Press.
———. 2010. “Language variation and complex systems”. American Speech 85: 263–86.
204 Lamont D. Antieau

——— and Susan Tamasi. 2003. “Distributional foundations for a theory of language change”.
World Englishes 22: 377–401.
Kurath, Hans and Raven McDavid, Jr. 1961. The Pronunciation of English in the Atlantic States.
Ann Arbor, MI: University of Michigan Press.
Pederson, Lee. 1990. Ms. “A plan for a Linguistic Atlas of the Western States”.
———. 1996a. “LAWCU project worksheets”. Journal of English Linguistics 24: 52–60.
———. 1996b. “LAMR / LAWS and the main chance”. Journal of English Linguistics 24: 234–49.
——— and Michael Madsen. 1989. “Linguistic geography in Wyoming”. Journal of English
Linguistics 22: 17–24.
Wagner, Susanne, and Liselotte Anderwald. 2007. “The Freiburg English Dialect Corpus
(FRED): applying corpus-linguistic research tools for the analysis of dialect data”. In Joan
C. Beal, Karen P. Corrigan and Hermann L. Moisl, eds. Creating and Digitizing Language
Corpora. Vol. 1. Synchronic Databases. Houndmills, New York: Palgrave-MacMillan, 35–53.
Zipf, George K. 1949. Human Behavior and the Principle of Least Effort. Cambridge, MA:
Addison Wesley.

Author’s address
Lamont Antieau
4021/2 Sixth Street
Racine, WI 53403
USA
lamont@antieau.org

You might also like