[Lingtyp] "grammatically encoded" - answer to Christian

Kasper Boye boye at hum.ku.dk
Thu Mar 9 13:58:52 UTC 2023

Dear Christian and all,

(Those of you who are not interested in the discussion of what it means to be grammatical, please forgive me and ignore this message).

As for stressability, Peter and I never claimed and would never claim that it is a criterion of lexical status. Rather, focusability is (and we were not the first to claim this). Thus, to the extent that stress is a reliable indicator of focus, it may serve as a criterion. But in many cases, it is not. For instance, when affixes are stressed, this (at least in many European languages) does not mean that they are focused independently of their stems.

As for your suggestion for an understanding of what it means to be grammatical (‘Such aspects of linguistic expressions are grammatical whose conformation obeys constraints of the particular linguistic system’), isn’t it circular? Unless the ‘linguistic system’ is grammar (and what else could it be?), it simply means: such aspects of linguistic expressions are grammatical whose conformation obeys constraints of the particular grammar.

A way out of this is to talk about grammatical dependencies based on conventionalized secondary status: just as second violins depend on first violins in relation to which they are secondary, grammatical elements depend on combination with host elements in relation to which they are secondary. This is why grammatical elements cannot constitute utterances on their own. In your 2013 review, you call this notion of dependency ‘obscure’, and maybe we didn’t explain it in enough detail in our contribution to the book you reviewed, but I think it is extremely clear and simple (even a couple of police officers from my football team got it).

With best wishes,

Fra: Lingtyp <lingtyp-bounces at listserv.linguistlist.org> På vegne af Christian Lehmann
Sendt: 8. marts 2023 19:46
Til: lingtyp at listserv.linguistlist.org
Emne: Re: [Lingtyp] "grammatically encoded"

Dear Kasper,

things are getting interesting here. Just a few observations:
Am 08.03.2023 um 17:03 schrieb Kasper Boye:

We claim in (Boye & Harder 2012) that this understanding of the lexical-grammatical distinctions is to a high extend co-extensive with traditional conceptions (otherwise, it would be an understanding of something else). As pointed out by Riccardo and Christian (in the 2013 review of the latter), however, it is true that in some cases, classifications based on our understanding are at odds with pretheoretical classifications. As mentioned by Riccardo and Christian, for instance, some pronouns come out as lexical. We believe that this is not a serious problem for the understanding we propose, because we believe our proposal offers a functional rationale for the status of grammar as a design feature of human languages. The fact that other features characteristic of grammatical elements are not fully co-extensive could also be regarded as a problem for the pretheoretical classifications. We would also like to point to the fact that the idea of distinguishing between lexical and grammatical pronouns is not entirely new and strange (cf. e.g. the distinction between weak and strong pronouns in Romance languages).

Lehmann op.cit. actually says:

While this may be an important observation, it requires us to deny grammatical status to stressable
formatives like interrogative, demonstrative and (tonic) personal pronouns, modal verbs,
negators and many others.

Thus, you would have to assume that all these stressable items have been regarded as grammatical erroneously by "pretheoretical classifications". They do, however, meet my definition, if to different extents.

It is also not seldom forgotten that grammatical affixes are not necessarily stressless (or clitic). Many Romance conjugation suffixes bear stress, and so do many Russian declension suffixes. You probably will not want to deny grammatical status to these. If so, you would have to sharpen the criterion of stressability.

As for Christian’s own suggestion for an understanding (‘Such aspects of linguistic expressions are grammatical whose conformation obeys constraints of the particular linguistic system’), we wouldn’t go as far as claiming to refute it, but would like to point out that it does not seem to distinguish lexical from grammatical elements: also lexical elements are constrained by the linguistic system – otherwise, we would not have distributional classes. In our view, structural constraints are what distinguishes both lexical and grammatical elements from holophrases.

Given grammaticalization, being governed by structural constraints is, again, a matter of degree. Constraints may add up on an item or class of items, and they may be more or less strict. If you have a distribution class for which there are rules of grammar, then that is - to that extent - a grammatical class. Nouns of a certain gender may be such a class; and gender certainly is a grammatical category. If you have a distribution class for which there are semantic rules, then that is a lexical class. Human nouns or agentive verbs may be such a distribution class. Moreover, this kind of class is generally not proper of a particular linguistic system.

Best wishes,


Prof. em. Dr. Christian Lehmann
Rudolfstr. 4
99092 Erfurt



christianw_lehmann at arcor.de<mailto:christianw_lehmann at arcor.de>



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lingtyp/attachments/20230309/9dcf5dd1/attachment.htm>

More information about the Lingtyp mailing list