Professional Documents
Culture Documents
As the study of language has evolved from the narrow confines of linguistic
form to include the study of language use, researchers have often found them-
selves either without the theoretical tools needed to pose and discuss the
issues or without a research methodology adequate to collect the appropriate
data.1 The study of deference and how it is associated with particular linguis-
tic structures appears to suffer from both difficulties. On the one hand, aside
from a few scattered comments in the recent linguistic and language-related
literature (cf. Goffman, 1971), there has been no account of what deference
is. Moreover, although speakers believe that they understood deference and
can recognize it, it is very clear that one cannot follow the linguistic tradition
and appeal directly to the intuitions of the native speaker to sort out the
degree of deference associated with particular expressions. To be sure, there
would be general agreement that the use of 'You ought to do that right now'
as a suggestion is far less deferential than suggest that you do that fairly
soon', but the use of such intuitions quickly breaks down on the more subtle
cases, and judges are inconsistent.
Nevertheless, this paper is an attempt to characterize deference and how it
appears to be systematically associated with certain types of linguistic forms.
We begin by presenting our view of deference, distinguishing it from polite-
ness with which it is often confused. We then report on two experiments
involving speakers of Spanish and English in which we employed a paired-
comparison methodology to determine both how deference was associated
with certain linguistic forms used for requesting as well as the similarity of
the associations across the two languages.
Deference
The experiments
Experiment I
We selected the act of requesting and, for the first experiment, 25 English
sentences which reflect frequently used requesting forms. These sentences are
This procedure resulted in some 2000 judged pairs from the 40 subjects. The
pairs were submitted to the program described above and a weighted rank-
order was obtained. The results are presented in Table 1.
Table 1.
For the scaling procedure adapted for this research, there is no fixed way
of interpreting the results. Of course, for a total range of 2.5, if one item is
scaled at 2.0 and another as 0.2, they reflect a meaningful difference on the
assumption that the entire procedure provides a meaningful scale. However,
when two items are closely scaled, any interpretation must be relatively sub-
Table 2.
most three ways in which one can differ from another: the form of the modal
is indicative versus conditional (can/could)', the modal is positive or negative
(can/can Y); the sentence is in the interrogative or imperative-tag form with
the modal can ('Can you do that'/'Do that, can you').6 We initially held a
very strong hypothesis concerning the role that these three linguistic features
would play in the assignment of deference to sentences in which they occurred:
first, that all interrogative forms would be judged more deferent than the
imperative forms; second, that, within each of the initial divisions, sentences
with a conditional modal would be judged more deferent than the indicative
cases; and third, that, within each of these four divisions, the sentence with
the positive modal would be more deferent than the negative one. This hy-
pothesis can be represented by Figure 1. The numbers in the parentheses in
the figure correspond to the ranking of the 8 sentences in Table 2 and there-
fore give an indication of the degree to which this strong hypothesis was
confirmed.
There are three clear results which emerge from considering these 8 sen-
tences: Whenever a pair of sentences differed in only one way:
1. A conditional sentence was judged more deferent than the corresponding
indicative one: (1/2; 3/7; 4/6; 5/8).
2. An interrogative sentence form was judged more deferent than the
corresponding imperative + tag form; (1/4; 2/6; 3/5; 7/8).
3. A sentence with the positive modal was judged to be more deferent than
the corresponding one with the negative modal; (1/3; 2/6;4/5; 6/8).
Several sentences were judged in ways which ran counter to the strong hy-
pothesis of deference ordering shown in Figure 1. The most flagrant violator
was the sentence 'Can't you do that' (negative-indicative-interrogative) which
was predicted to be in fourth place but slipped to seventh, a clear seventh.
-Positive (4)
Conditional^
-Negative (5)
Imperative + Tag:
-Positive (6)
^Indicative =
-Negative (8)
Figure 1.
This caused the claim that all interrogatives would be judged more deferent
than imperative + tag cases to be discontinued. In addition, the 'Can you do
that' and 'Couldn't you do that' were reversed in the predicted rank, though
the difference here is relatively small. In general, we see the results of these
8 sentences to strongly corroborate the general hypothesis that particular
linguistic forms do contribute substantively and systematically to the assign-
ment of deference to a sentence.
The same phenomenon, involving the modal wz//, showed mixed results. On
the corroborative side, 'Will you do that' (2) was judged to be more deferent
than it's negative version 'Won't you do that' (9), the imperative 4- tag versions
'Do that, will you' (18) and 'Do that, won't you' (17). In addition, 'Would you
do that' (3) was judged to be more deferent than 'Do that, would you' (14)
and this latter more deferent than 'Do that, won't you' (17). However, the
'will you' form (2) was seen to be slightly more deferential than the 'would
you' case (3) and the 'Do that won't you' (17) more than the 'Do that, will
you' (18), though here again, a scale difference of 0.14 or so should not be
taken to be meaningful.
Our initial feeling was that the will cases in whatever version would be
judged to be more deferential than the corresponding can sentences. This was
confirmed. 'Will you do that' (2) was rated over 'Can you do that' (8) by a
sizeable margin though the 'would you' form was judged only slightly more
deferent than the 'could you' version. 'Won't you do that' (9) was rated well
above 'Can't you do that' (22) as were the 'Do that, will you' and 'Do that,
Experiment II
Having confirmed our initial hypothesis that English speakers could system-
atically assign a relative level of deference to sentences used to request, we
turned to the question of whether speakers from another language would mirror
the English-speaking judges. Fräser (1979) argued that the strategies used to
perform a particular speech act are essentially the same across languages. For
example, if 'can you... .' can count as a request in English, the semantically
equivalent sentence in Spanish, Twi or Chinese will be used as a request as
well. What was of particular interest, however, was the degree to which the
perceived relative degree of deference of the English sentences would be
mirrored by other language groups.
Subjects for this experiment were 25 college students in the Boston area
who spoke Spanish as their first language. The sentences chosen were 14 of
the 25 used by the English speakers, chosen to reflect the full range of the
English results. The same manner of presentation of the stimulus sentences
was used as earlier, and the same analysis was performed. In addition, the
original English data was reanalyzed for only the 14 English sentence stimuli
pairings corresponding to the Spanish sentences. These results are presented
side-by-side in Table 3, with the number in parenthesis in the Spanish data
indicating the corresponding English sentence.
Several things are clear from these results. First, there is very close agree-
ment on the rank ordering of the 14 sentences. Leaving aside Spanish sentences
numbers (3), (10) and (6) which fail to correspond closely to the ranking of
the corresponding English sentence, the revised rank ordering is the following:
E: l 2 3 4 5 6 7 8 9 10 11
S: 2 1 4 3 5 8 6 7 9 10 11
For example, 'Do that' and 'You have to do that' were ranked the least defer-
ential in English; they were ranked similarly in Spanish. At the other end of
the spectrum, 'Could you do that' and 'I'd appreciate it if you'd do that' were
ranked the most deferential in both languages. Very little difference in ranking
was found in between. It was impossible to reanalyze the data with only 11
examples because the number of English examples, already reduced from 25
sentences to 14, did not provide sufficient numbers for each pairing of the
11 sentences with the other 10 to insure reliable results.
What about the three maverick sentences? The first, Spanish (3) ('Necesita
hacerse') was judged far less deferential than its English counterpart ('That
needs to be done'). Our best estimate for this difference lies in the translation
of the English which might better have been 'Hay que hacerlo' or perhaps
'Se necesita hacerlo9 since the '/o' of Spanish which though often translated as
Boston University
Notes
We would like to express our appreciation to both Jean Fräser and Ellen Rinteil for
their insightful and constructive comments on earlier versions of this paper. Their
lack of deference in tendering their thoughts made the revising go more quickly.
References
Austin, J.L. (1962). How to Do Things with Words. Cambridge, Massachusetts: Harvard
University Press.
Fräser, B. (1975). Hedged performatives. In Syntax and Semantics, vol. 3, Speech Acts,
P. Cole and J. Morgan (eds.). New York: Academic Press.
- (1979). On requesting: A study in pragmatics. Unpublished manuscript, Boston
University.
- (in press). Pragmatic research in second language acquisition.