[Lingtyp] Moods and non-finites?

Sat Jul 15 10:22:18 UTC 2023

Dear Bjoern – Just a couple quick rejoinders:

  *   Yes, I do view deontic modality as operating on propositions, in accordance with the semantic and philosophical literature that I’m familiar with (but please note I’m by no means an expert on modality). Deontic modality is in this sense different from descriptions of ability, particularly generic statements such as your example (3). But of course there is pragmatic overlap here. Ernie is able to eat 20 pancakes in one sitting and Ernie eats 20 pancakes in one sitting are semantically distinct, but they may be used to answer the same question under discussion by conveying the same nugget of information.
  *   I also understand, and am sympathetic to, the reluctance to treat Ernie is able to eat 20 pancakes in one sitting as a modalization of the proposition ‘There exists, or will exist, an event during which Ernie eats 20 pancakes in one sitting.’ Language is “organic”, theories are not, or not as organic as natural languages. Just because we can come up with a unified account of the semantics of modality doesn’t mean that all expressions covered by this account fit equally well with our intuitions of how they are interpreted or how they evolve.
  *   I’m a little bit worried about your examples (1a) and (2a), because those are in my view not merely modalized statements. The matrix predicate there seems to express a particular kind of speech act that “colors” the modal interpretation.
  *   Mood operating on speech acts is not inconsistent with its being marked in subordinate clauses etc. Subordinate clauses usually do not have their own topic situations (exceptions arise with apositives and parentheticals), but rather are interpreted with respect to the topic situation of the superordinate utterance. Since you mention Micky Noonan’s (fantastic!) work, this is one place where his notion of ‘dependent time reference’ comes in. For examples, subjunctives are often used to describe unrealized situations that are desired or feared (etc.) to occur at some unspecified distance into the future of the matrix utterance’s topic situation.

HTH! – Juergen

Juergen Bohnemeyer (He/Him)
Professor, Department of Linguistics
University at Buffalo

Office: 642 Baldy Hall, UB North Campus
Mailing address: 609 Baldy Hall, Buffalo, NY 14260
Phone: (716) 645 0127
Fax: (716) 645 3825
Email: jb77 at buffalo.edu<mailto:jb77 at buffalo.edu>
Web: http://www.acsu.buffalo.edu/~jb77/

Office hours Tu/Th 3:30-4:30pm in 642 Baldy or via Zoom (Meeting ID 585 520 2411; Passcode Hoorheh)

There’s A Crack In Everything - That’s How The Light Gets In
(Leonard Cohen)
--

From: Wiemer, Bjoern <wiemerb at uni-mainz.de>
Date: Saturday, July 15, 2023 at 11:41
To: Juergen Bohnemeyer <jb77 at buffalo.edu>, Jussi Ylikoski <jussi.ylikoski at utu.fi>, lingtyp at listserv.linguistlist.org <lingtyp at listserv.linguistlist.org>
Subject: AW: Moods and non-finites?
Dear All,
if I my add some questions to this interesting discussion, I’d first like to say that I’m much sympathetic with Jürgen’s attempt at drawing apart “mood” and “modality”. I could even “buy” his definitions – as a working basis – but I find it important to specify some things.
            First, when modality is generally defined as a domain in which NEC(essity) and POSS(ibility) operate on _propositions_, this raises the question what you do with non-epistemic modality. While epistemic judgments clearly reside on propositions, deontic and other kinds of modality on “lower layers” are not related to propositions, but to states of affairs – unless you want to say that, e.g., a moral judgment like

(1)       People should not kill

or prohibitions like

(2)       You are not to smoke in this room

or characterizations that can easily be read like assignments of general dispositions, as in

(3)       Ernie eats 20 pancakes in one take

are seen as “translations” of something like

(1a)      It is required that people should not kill.
(2a)      We urge you not to smoke in this room.
(3a)      Ernie is capable of eating 20 pancakes in one take.

(Note that 1-3 are not truth-conditionally falsifiable.)
I mean there is a difference between uttering a judgment that is not epistemic (an obligation, a statement about somebody’s dispositions, and more) and giving a “description” of that non-epistemic judgment. Otherwise I don’t understand how we might say that ALL modality is about NEC:POSS-contrasts over propositions.

Second, if “mood” is defined as relating to illocutionary force, then this rules out mood distinctions that have been made for subordinate clauses. Somewhat ironically, the term “subjunctive”, or “conditional”, seems to have been coined for means (morphological markers on the verb or the VP) that occurs only, or predominantly, in subordinate clauses; and subordinate clauses are often claimed to be void of independent illocutionary force.

In addition, above all that has been said on these intriguing issues so far since Jussi came up with his question, the question of “mood” becomes complicated even more when we take into account that in certain, hm, traditions of linguistic research “mood”, or more specific terms like “subjunctive”, have been applied to scattered marking on clause level that likewise distinguishes things like illocutionary force (representative vs directive speech acts) or that occur in certain control constructions (I want SoA, I order that SoA, etc.), for clausal complements with weakened epistemic support (e.g., after negated cognitive or perception verbs: I don’t think/believe that P; I don’t see that P), or in other structures of clause combining that are in some languages (e.g., Romance) regularly marked with a morphological subjunctive/conditional, or in other (recent) traditions are discussed under headings like “irrealis” (understood as a notional distinction, not a class of tight structural means).
In many cases, “analytical subjunctive” has been referred to structures as in Balkan Slavic, where all these aforementioned notional distinctions and structural environments are marked (or can be marked) with no specific morphology (on the verb or other constituents of the clause), but by clause-initial connectives (the proper status of which has been subject to countless debates) that contrast with standard (neutral, default) complementizers (Engl. ‘that’). See (4a) vs (4b) from Bulgarian: the standard complementizer če (associated with full epistemic support) in (4a) can occur with any tense (here: the aorist, rešix), while the connective da occurs, among other things, in complements of predicates that code an intention, as in (4b); da restricts the array of tenses which are otherwise allowed in the “indicative” (da cannot occur with the aorist, but preferably occurs with present tense or the perfect).

(4a)        Mislj-a, če reši-x vsički zadači
think[ipfv]-prs.1sg comp solve[pfv]-aor.1sg all task.pl
na testa.
on test.def
‘I think I (have) solved all tasks from the test.’

(4b)        Mislj-a da izlezn-a malko na văzdux.
think[ipfv]-prs.1sg con go_out[pfv]-prs.1sg a_bit on air
‘I think I’ll go a little bit out into fresh air.’

There is, thus, a clear contrast between če and da as means of connecting clausal complements to higher-order predicates, and da is also used in independent clauses, e.g. as kind of imperative (e.g., Macedonian Da gi prečekate! ‘Welcome them!’). Both connectives are practically in complementary distribution, and they differ as for the range of verb forms with which they can combine, but there is no special verb morphology designed to mark all the aforementioned clause types.
In the literature, da-clauses are often dubbed “analytic subjunctives” (or simply “analytical mood”). This parlance came up, as far as I can see, predominantly in generative literature, but it has been adopted in some (prominent) linguistic work, such as Noonan’s article on clausal complementation. Though I personally find this unfortunate (it creates more confusion than it helps clarifying how languages are structured, and how they are structured differently), this is just a fact of linguistic practice. The least one can (and should) do, in my view, is to clearly say whether “mood” (or any assumed subcategory, like “subjunctive”) is to be defined as a language-specific structure (e.g., the traditional subjunctive marked on the verb) or as a notional domain. If the latter is at stake, then we nonetheless need clear-cut form/structure-oriented criteria saying whether illocutionary force distinctions or distinctions like full vs weak epistemic support in clausal complements, or of what happens in control constructions (e.g., after manipulative verbs), are marked by clausal connectives (a prime means in Slavic languages), or by verbal morphology (as probably in Romance) or by both (as in Romanian, unsurprisingly as an “overlap” between Slavic and Romance). Otherwise things like “mood choice” and “complementizer choice” will inevitably run into one another and become indistinguishable, at least if one looks out for comparative concepts.

Sorry for bothering you. But the intricacies, and the pervasive relevance of these considerations can be seen in Slavic languages par excellence. I happened to bring this down in a (hopefully) comprehensive way in a recent paper available under
            https://doi.org/10.1515/slaw-2023-0012
or on
https://uni-mainz.academia.edu/Bj%C3%B6rnWiemer
(Between analytical mood and clause-initial particles – on the diagnostics of subordination for (emergent) complementizers. Zeitschrift für Slawistik 68-2, 187-260)

I have no solution for how to define a comparative concept of “mood”, but it seems obvious that form- and function-oriented criteria must not be used interchangeably (to say the least).

Best,
Björn (Wiemer).

Von: Lingtyp [mailto:lingtyp-bounces at listserv.linguistlist.org] Im Auftrag von Juergen Bohnemeyer
Gesendet: Freitag, 14. Juli 2023 22:32
An: Jussi Ylikoski <jussi.ylikoski at utu.fi>; lingtyp at listserv.linguistlist.org
Betreff: Re: [Lingtyp] Moods and non-finites?

Dear Jussi – I see that you’re getting as many opinions as you’re getting responses 😊 But I’m going to add mine, simply because none of the other users seems to have mentioned what I see as the primary issue.

Fwiw., from my point of view, the key question here is not finiteness, but semantics. Let’s assume that _-ne_ does indeed have a single “semantic meaning” (as opposed to pragmatic implicatures arising from it in various contexts), and that meaning is, as you say, epistemic modality.

The question then arises whether one wishes to conflate the notional categories (a.k.a. comparative concepts) of ‘mood’ and ‘modality’ or to keep them distinct. Personally, I prefer the latter, since they are quite distinct conceptually and in my experience, their expression tends to be in first approximation independent, or at least distinct, in the languages of the world. But, again, this is a consideration purely of utility – there’s no correct or incorrect answer.

If I were to propose typologically useful working definitions for the comparative concepts ‘mood’ and ‘modality’, I would offer something like this:

  *   Modality classifies propositions into those that are necessarily (not) true and those that are potentially (not) true. (Non-modalized speech acts simply steer clear of this classification.) This distinction can be conceptualized in terms of quantification over possible worlds, force dynamics, or however else one sees fit. And to say that a proposition is necessarily or possibly true requires some, usually contextually determined, set of assumptions against which the modalized speech act is evaluated. In the case of epistemic modality, that set of assumptions is the sum of what the speaker purports to know.

  *   Mood classifies utterances in terms of whether their ‘topic situation’ is (assumed to be) part of the interlocutors’ actual world (the speech situation and its past) or not (the speech situation’s future and counterfactual past situations). An utterance’s topic situation is the situation it makes a statement or promise, asks a question, issues a directive, etc., about.

Now, I’m not aware of any theory of finiteness I’d be ready to endorse, including my own. (I guess you could take this this along the lines of the famous Groucho Marx quip that he’d never join a club that would accept him as a member.) But, based on the above characterization, modality is a propositional operator, whereas mood is a speech-act-level operator. And since propositions are the objects/arguments of speech acts, we should expect to generally find modality further “down” in the syntactic structure, i.e., closer to the lexical content, than mood.

So this means that if _-ne_ is indeed a modal operator, I would consider its showing up in nonfinite projections less surprising than if it were a mood operator.

Lastly, none of the above should be construed as a criticism of the traditional terminological practice of Finnish linguistics. The cognancy of _modal_ and _modus_ alone makes it almost unreasonable to expect terminological traditions to be answerable to the semantic factoids cited above. Particularly when paired with paucity of thorough and typologically sensitive semantic work on the issues.

HTH! – Juergen

Juergen Bohnemeyer (He/Him)
Professor, Department of Linguistics
University at Buffalo

Office: 642 Baldy Hall, UB North Campus
Mailing address: 609 Baldy Hall, Buffalo, NY 14260
Phone: (716) 645 0127
Fax: (716) 645 3825
Email: jb77 at buffalo.edu<mailto:jb77 at buffalo.edu>
Web: http://www.acsu.buffalo.edu/~jb77/

Office hours Tu/Th 3:30-4:30pm in 642 Baldy or via Zoom (Meeting ID 585 520 2411; Passcode Hoorheh)

There’s A Crack In Everything - That’s How The Light Gets In
(Leonard Cohen)
--

From: Lingtyp <lingtyp-bounces at listserv.linguistlist.org<mailto:lingtyp-bounces at listserv.linguistlist.org>> on behalf of Jussi Ylikoski <jussi.ylikoski at utu.fi<mailto:jussi.ylikoski at utu.fi>>
Date: Thursday, July 13, 2023 at 20:22
To: lingtyp at listserv.linguistlist.org<mailto:lingtyp at listserv.linguistlist.org> <lingtyp at listserv.linguistlist.org<mailto:lingtyp at listserv.linguistlist.org>>
Subject: [Lingtyp] Moods and non-finites?
Dear typologists,

I have a question about the notion of mood as a grammatical, and more specifically, morphological category. It should not have come as a surprise to me that on a general, language-specific level, various dictionaries of linguistic terms define and characterize mood quite vaguely, but I am still surprised to see how difficult it is to find explicit statements on whether or not morphological moods are generally limited to finite verb forms or not.

Put concretely, I am wondering whether it is conceptually (or typologically) odd or natural to regard the Finnish "potential mood" marker -ne- a mood, as its use is not strictly limited to finite verb forms (2) but can also attested in some – and only some – non-finites such as the present participle (2), in contrast to the unmarked or "indicative" participle seen in (3). The semantic function of the potential in -ne- is that of epistemic modality:

(1) Remontti   valmistu-ne-e        elokuu-ssa.
    renovation be.completed-POT-3SG August-INE
    'The renovation will probably be completed in August.'

(2) elokuu-ssa valmistu-ne-va            remontti
    August-INE be.completed-POT-PTCP.PRS renovation
    'the renovation that will probably be completed in August'

(3) elokuu-ssa valmistu-va           remontti
    August-INE be.completed-PTCP.PRS renovation
    'the renovation that will be completed in August'

In light of the Finnish (and more generally Uralic) grammatical tradition, participials seen in (2) do not and cannot exist, but if they do as it seems, inflectional moods do not behave like this, and the morpheme -ne- in valmistu-ne-va should probably be analyzed as a kind of derivational affix instead.

I would be interested to know whether linguists outside the Finnish tradition see it as problematic or unproblematic to call the morpheme -ne- in valmistu-ne-va (2) a participle a marker of a morphological (inflectional) mood. Are there any parallels to forms like this, and possibly studies on these issues?

Best regards,

Jussi

https://users.utu.fi/jumyli/

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lingtyp/attachments/20230715/9d0d9354/attachment-0001.htm>