Unidirectionality?

[Alan R. King]

Ron Kuzar asks about unidirectional diachronic developments such as "angers
me" > "I am angry", and the larger implications of such facts.  Concerning
such expressions of feeling, I have two comments to make:

(a) One way of viewing the development that RK mentions might be in terms of
its implications about the concept "subject": perhaps we may say that in a
language that uses "I am angry" the idea (read perhaps: range of semantic
roles) associated with the grammatical notion "subject" is ever so slightly
(or even not so slightly?) different from the idea in one that uses "angers
me".  So the diachronic development involves a partial redefinition of
"subject" in the language.  This hypothesis would suggest the likelihood
that in languages where the change mentioned takes place, this will not be
an isolated phenomenon but part of some larger process.

(b) Subject status can also have certain pragmatic implications.  Thus there
is a widespread strategy of what Brown & Levinson called "negative
politeness" consisting of avoiding giving the real protagonist (semantic
subject) the formal status of subject.  This can be achieved by complete
impersonalization, e.g. in French (and similarly in Hebrew, I think) saying

        Il faut regarder le livre.
        "It is necessary to look at the book."

when one actually means

        Tu dois regarder le livre.
        "You must look at the book."

or by an intermediate solution such as:

        Il faut que tu regardes le livre.
        "It is necessary that you look at the book."

where we have at least avoided "Tu dois...", which, with the semantic
subject as grammatical subject of the modal verb, is felt to be more direct
(and so, in certain communicative contexts, potentially less polite).  In
some languages, another alternative is to express the semantic subject as a
grammatical dative:

        Il te faut regarder le livre.
        "It is necessary to you to look at the book."

The flip side of this is that, since speakers *know* what is going on here,
there is a sense in which all the above constructions are merely different
grammatical expressions of a single pragmatic structure in which "you" is in
a single pragmatic relationship to the modalized predicate "must look",
corresponding to what I have called semantic subject.

Now, can we think of pragmatic reasons why it might be desirable to express
the semantic subject of "expressions of feeling" in a less direct manner?
Of course we can.  Such expressions can easily constitute what have been
called "face-threatening acts" in politeness theory, thus motivating
compensating politeness strategies.  On the other hand, if a language recurs
to a strategy such as dativizing a certain type of semantic subject
constantly, this can become conventionalized to the point where the original
politeness strategy is no longer clearly perceived by speakers.  At this
point, a plausible next step is the language's evolution is to "regularize"
the system by encoding the perceived semantic subject as grammatical subject....

The way in which I think the scenario just described is possibly relevant to
RK's (broader) question is as follows.  In cases where a change is found (or
thought) to be unidirectional, meaning that we find A>B but not *B>A, there
may nevertheless occur B>C and C>A, and therefore B>C>A, which is an
indirect form of B>A.  The problems are to identify C and to understand the
nature of the processes leading (a) to, and (b) from C.  What I have just
tried to do is suggest a possible identification and explanation for the
case of "expressions of feeling".  Since I have posed it in speculative and
non-empirical terms, others might be interested in providing illustrative
diachronic examples, if there are any to be found.


Alan R. King          |  EMAIL: mccay at jet.es
Indamendi 13, 7C      |  [or if all else fails] 70244.1674 at compuserve.com
20800 Zarautz         |  FAX: +34-43-130396
Gipuzkoa
Euskal Herria / Basque Country (Spain)