[Lingtyp] Definition of "personal pronoun"

[Sebastian Löbner]

Dear Juergen and everybody,

I agree that my condition 4 seems to be contradicting condition 6. This 
is due to an unfortunate formulation: I meant the ‘or’ non-exclusive. 
Your formulation is better.

As to the distinction between 1pp and 2pp (first/second person personal 
pronouns) on the one hand and 3pp on the other, I agree with Haspelmath 
that they should be treated separately. 3pp are functionally very 
similar to distal demonstratives, and in many languages formally 
identical. They have nothing to do with SAP (speech act participants), 
except that they cannot refer to speaker or addressee: This is the only 
connection. There is something like a referential ranking: 12pp cannot 
include reference to the speaker, and the referene of 3pp excludes 
speaker and addressee.

Let S be short for speaker/producer, A for addressee, P for any other 
person, X for non-person referents; * means one or more of the same:
2pp cannot include S in their reference, and 3pp neither S nor A.

- 1pp singular refer to S

- 1pp non-singular:  S + A*, S + A*+ P*, S + P*, maybe also S*.

- 2pp singular: A

- 2pp non-singular: A* or A* + P*, but never including reference to S.

- 3pp singular: P or X

- 3pp non-singular: P* or X* (mixed reference to persons and non-persons 
seems to be excluded)

As to the sparseness of sortal content: Yes, there are lexical nouns 
with very general content, but these are sortal nouns, denoting some 
SORT of thing. For example, English /he/ and /man/ have essentially the 
same sortal content “male person”, but /he/ is inherently unique, or 
definite, meaning something like “THE male person”, while /man/ is a 
sortal noun, meaning “(A) male person”. Inherently unique lexical nouns 
(often called “uniques”) include cases like “sun”, “moon”, “queen” etc. 
They denote any kind of unique institution in the world or in a smaller 
context. They have pretty specific content.
In the case of 3pp, which are uniques, too, it’s the lack of sortal 
content that makes them the “joker” expressions for heavily 
context-dependent unique reference. Thus, I think, my conditions 
properly rule out a confusion of pp, even 3pp, with ordinary nouns. In 
addition, there are the observations that pp differ in not admitting 
determiners or restrictive attributes; they are just atomic definite NPs.

Still, there is more reason for differentiation:

- 1pp only are restricted to persons, because they refer to something 
able of maintaining the role of a speaker in verbal communication 
(actually, there is more involved than mere production of speech signs).

- The reference of 2pp is less restricted. For example, people talk 
their cats and dogs and this seems to be a regular use of 2pp at least 
in some languages.

- If we consider 3pp as demonstratives, they may still differ in lacking 
deictic differentiation and they may be restricted to reference to 
persons. Japanese has 3pp /kanojo/ “she” and /kare/ “he”. These were 
introduced in the Meiji era for the translation of 3pp in European 
languages. Unlike Japanese 1pp and 2pp, they do not carry any social 
meaning, and unlike other Japanese demonstratives they lack deictic 
differentiation and refer exclusively to persons. The functional 
equivalent(s) in Japanese of /it/ are real demonstratives.

Best, Sebastian L

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lingtyp/attachments/20210714/f27fc22c/attachment.htm>