Languages with mood but not aspect or tense?

[Mike Yams]

Are there any languages which obligatorily mark mood, but do not mark, or optionally mark, tense and/or aspect?

[Richard W]

Thai feels like a good candidate. The marking of irrealis feels obligatory, whereas aspect and tense seem to be optional, especially if there is some other indicator of when an event happened or will happen. Now the irrealis marker, /dʑà/ (จะ), is often translated into English as 'will', so my claim would certainly be challenged on other forums.

[Salmoneus]

English, maybe?

Tense marking and to a lesser extent aspect marking are optional thanks to the use of 'narrative' presents and perfectives [although note that this isn't a true narrative screeve, because both perfective and imperfective aspects can be used with it]
As dawn breaks, the streets are empty; six hours later, there is carnage.
A man walks into a bar in the middle of the 20th century
So then he says to me 'blah', and I turn round to him and I say 'blah', and then he's like 'blah', and I'm like 'blah', and it is so weird...

But you never see this happen with modals. "I eat the fish" can never mean "I can eat the fish" or "I should eat the fish". Or maybe it can? We do have modal adverbs, so maybe "Potentially, I eat the fish" or "Ideally, I eat the fish" count... but then they don't line up nicely with the modals themselves, so some forms of modality are probably obligatory. I can't find a way to use the realis in place of "I should eat the fish," for instance. Or even really "I can eat the fish" - "feasibly, I eat the fish" feels neither a natural thing to say nor exactly the same. Of course, it could be rephrased with an infinitive and a noun: "I have the capability to eat the fish".

I guess partly it depends on what is meant by 'obligatory', since it's common for there to exist some alternative construction to get around most 'obligatory' things...

[Mike Yams]

That's a good point. I guess because in English (and I'd assume most languages) indicative is unmarked, it's hard to ever get rid of it.

[zompist]

Tense marking and to a lesser extent aspect marking are optional thanks to the use of 'narrative' presents and perfectives [although note that this isn't a true narrative screeve, because both perfective and imperfective aspects can be used with it]
As dawn breaks, the streets are empty; six hours later, there is carnage.
A man walks into a bar in the middle of the 20th century
So then he says to me 'blah', and I turn round to him and I say 'blah', and then he's like 'blah', and I'm like 'blah', and it is so weird...]

Morphologically, every verb here has an obligatory marked tense. We call it the present, but of course it's not hard to find examples like yours when it's not used semantically for the present time. We also use it for future actions: "Tomorrow he goes to Berlin." (Similarly, we use the past tense as a hypothetical: "If I went to Berlin, I'd visit Kreuzberg.")

I expect most languages have conventional exceptions for the use of various verb forms. E.g. French often narrates past actions in the present or, weirdly, the future, and uses the imperfect for suggestions ("Si nous allions...")

But you never see this happen with modals. "I eat the fish" can never mean "I can eat the fish" or "I should eat the fish".

It's hard to have the simple present represent a specific modal, but general irrealis is easy:

"I hope he dies of an embolism."
"If I fell in love with you, would you promise to be true?"
"We sell this thing and we live high on the hog for a year."

To go back to the OP, my understanding (from Palmer) is that the irrealis almost always coexists with tense or aspect. One language where it doesn't is Maram (a language of India); I have examples in ALC p. 155.

[Travis B.]

One thing is that the hypothetical use of the past in English likely is a continuation of the English past subjunctive, which largely disappeared not through its being replaced altogether by a different form but by becoming indistinguishable from the past indicative aside from, in some varieties such as most NAE varieties, were and had, and its use by itself becoming highly archaic aside from at least the words were, had, would, should, and could and in some varieties many of those as well. Hence one could argue that such "past indicatives" are really past subjunctives.

[Mike Yams]

Is the English past subjunctive like the present subjunctive in that it's used every now and then, if very rarely, or is it completely gone? (I can't actually think of an example of English past subjunctive off the top of my head.)

[zompist]

One thing is that the hypothetical use of the past in English likely is a continuation of the English past subjunctive, which largely disappeared not through its being replaced altogether by a different form but by becoming indistinguishable from the past indicative aside from, in some varieties such as most NAE varieties, were and had, and its use by itself becoming highly archaic aside from at least the words were, had, would, should, and could and in some varieties many of those as well. Hence one could argue that such "past indicatives" are really past subjunctives.

One could, but it would be silly. English doesn't need to have a subjunctive just because French and Latin do, or even because it once did.

Picture a child learning English. She does not hear or see anything which distinguishes the "subjunctive" from the past. She doesn't learn historical forms which might have differed.

There's only one anomaly, and it's not even a separate form: the construction "(if) I were". But this is better explained as an odd quirk of 'to be' (one of several) rather than a different verb form that suddenly appears on the thousands of other verbs in English. Or even as a curious literary retention, like "methinks" (which does not prove that we have a dative).

If you really think that one anomalous form of 'to be' merits adding a "subjunctive" to every verb in English, do you do the same for the two past participles of "hang"?

[Travis B.]

Is the English past subjunctive like the present subjunctive in that it's used every now and then, if very rarely, or is it completely gone? (I can't actually think of an example of English past subjunctive off the top of my head.)

The past subjunctive is only distinguished from the past indicative for be (as were) Otherwise the two have phonetically merged completely. However the old usages of the past subjunctive have been retained and are still used today - they are just in most cases just not distinguished from the past indicative.

[KathTheDragon]

I think the point to be made here is that form and function should not be confused. Formally, English has no subjunctive, except for a few irregular forms. Functionally, however, the subjunctive is alive and kicking, a good example being the hypothetical past.

[zompist]

Is the English past subjunctive like the present subjunctive in that it's used every now and then, if very rarely, or is it completely gone? (I can't actually think of an example of English past subjunctive off the top of my head.)

The past subjunctive is only distinguished from the past indicative for be and have (as were and had). Otherwise the two have phonetically merged completely. However the old usages of the past subjunctive have been retained and are still used today - they are just in most cases just not distinguished from the past indicative.

The past indicative of "have" is "had"; it's no exception to the merger.

[Travis B.]

Is the English past subjunctive like the present subjunctive in that it's used every now and then, if very rarely, or is it completely gone? (I can't actually think of an example of English past subjunctive off the top of my head.)

The past subjunctive is only distinguished from the past indicative for be and have (as were and had). Otherwise the two have phonetically merged completely. However the old usages of the past subjunctive have been retained and are still used today - they are just in most cases just not distinguished from the past indicative.

The past indicative of "have" is "had"; it's no exception to the merger.

I realized that after I wrote the post but before I could get back to my machine.

[Travis B.]

I think the point to be made here is that form and function should not be confused. Formally, English has no subjunctive, except for a few irregular forms. Functionally, however, the subjunctive is alive and kicking, a good example being the hypothetical past.

You mean morphologically English has no past subjunctive as distinguished from the past indicative, except for were for be, but functionally it still has a past subjunctive, e.g. the hypothetical past.

Morphologically at least most of NAE still has a present subjunctive for all verbs, as marked by the lack of -(e)s for the 3rd sg., which is still used in a range of subordinate clauses.

[KathTheDragon]

Yeah, that.

[So Haleza Grise]

Yukulta has no perfect/imperfect distinction: Mood is marked, and tense in a more limited way, but the primary distinction is between realis and irrealis. Sandra Keen argues that the realis/irrealis distinction functions as underlying perfect/imperfect, but it's quite indirect.

[cntrational]

There's only one anomaly, and it's not even a separate form: the construction "(if) I were". But this is better explained as an odd quirk of 'to be' (one of several) rather than a different verb form that suddenly appears on the thousands of other verbs in English. Or even as a curious literary retention, like "methinks" (which does not prove that we have a dative).

If you really think that one anomalous form of 'to be' merits adding a "subjunctive" to every verb in English, do you do the same for the two past participles of "hang"?

My own dialect of English, and many others I've seen, has been systematically replacing the subjunctive were with was. So it's "If I was..." to me.

[Richard W]

If you really think that one anomalous form of 'to be' merits adding a "subjunctive" to every verb in English, do you do the same for the two past participles of "hang"?

Hang, hanged, hanged and hang, hung, hung are two different verbs (the former meaning 'to hang by the neck until dead'), or the past forms are in free variation, or some mixture of the two. Invoking mood makes no sense.

[zompist]

If you really think that one anomalous form of 'to be' merits adding a "subjunctive" to every verb in English, do you do the same for the two past participles of "hang"?

Hang, hanged, hanged and hang, hung, hung are two different verbs (the former meaning 'to hang by the neck until dead'), or the past forms are in free variation, or some mixture of the two. Invoking mood makes no sense.

I wasn't. It's an analogy.

* 'Hang' has two past forms. Yet we don't use that as a reason to create an extra category of inflection for every other verb; we just say 'hang' is a strange verb.
* 'Be' is anomalous in that it has 'I was' vs. '(if) I were'. To be consistent, we shouldn't invent a 'subjunctive' that every other verb has.

[Richard W]

* 'Be' is anomalous in that it has 'I was' vs. '(if) I were'. To be consistent, we shouldn't invent a 'subjunctive' that every other verb has.

Actually, it does create a subjunctive for almost every other verb, as in the contrast of, 'This bed was slept in last night' and 'If this bed were slept in last night...'. I had to think hard to find a verb that didn't have a 'be' passive form in my idiolect - I think 'become' doesn't have a passive form. I'm not sure about 'be' - is 'Many famous characters had been been by that actor' grammatical in anyone else's 'Standard English' idiolect?

Now, you may prefer to argue that the contrast of a simple past subjunctive passive and a simple past indicative passive shouldn't be imported into the active voice. After all, who would seriously derive passives from actives?

[zompist]

* 'Be' is anomalous in that it has 'I was' vs. '(if) I were'. To be consistent, we shouldn't invent a 'subjunctive' that every other verb has.

Actually, it does create a subjunctive for almost every other verb, as in the contrast of, 'This bed was slept in last night' and 'If this bed were slept in last night...'.

Um, that's 'be'. This alleged subjunctive has no separate forms except for 'be', that's what I'm pointing out.

[Salmoneus]

But that's sophistry, because we all know that we use a periphrastic conjugation for the passive.

"The bed was slept in last night" = 'be' as a verb if and only if you are talking about a state, 'being "slept in"', that the bed temporarily had last night. But in reality, we 99.9% of the time we are using it as the past passive of 'sleep'. It's like arguing that no language has a subjunctive because really it's only the affix that has a subjunctive form, not the verb...


But then, this whole debate is sophistry. We know there is (obviously) a subjunctive mood in English. Saying "oh, but specifically in the past tense, for most verbs, this mood isn't morphologically marked, therefore for specifically the past tense it doesn't exist" is on a level with "[h] is an allophone of /N/". Or, indeed, the equivalent "Latin didn't have a present passive imperative, it just used the infinitive", or "Latin didn't agree with person, mood or tense in the passive perfect, except for the verb 'esse'". You CAN say these things, it just has no useful meaning or purpose. It's playing with words.

[Richard W]

And, of course, "English has no present subjunctive, it just uses the imperative instead".

Actually that seems to be true for my idiolect for indirect commands as 'the requirement that it deliver 20,000 gallons an hour'. Such clauses resist the rule of sequence of tenses! However, for unlikely protases I do have a 3s present suffix, something like /∅ː/, which has the same duration as /s/. It has a lot of very distinct allophones, and I suspect people I'm talking to don't notice the difference from ///∅///.

[Richard W]

We know there is (obviously) a subjunctive mood in English. Saying "oh, but specifically in the past tense, for most verbs, this mood isn't morphologically marked, therefore for specifically the past tense it doesn't exist" is on a level with "[h] is an allophone of /N/".

Actually, a lot of British people don't have a separate subjunctive, but use the indicative or a periphrastic mood in indirect commands, and completely lack the subtlety of a present subjunctive for use with 'if'. The lack of a past subjunctive wouldn't be a problem, but would be comparable with Latin lacking a future subjunctive. However, some of us do have a past subjunctive, and I've an uncomfortable feeling I pronounce it slightly differently to the past indicative. The differences don't match any theory I'm aware of.

[cntrational]

English has a subjunctive, it's just morphosyntactically marked, not pure affixal morphology.

There's no reason to say that English doesn't have a subjunctive because it's marked by auxilliary verbs or periphrastic constructions. The emphasis on affixal morphology alone is a holdover from Greco-Latin grammarians; syntax is just as important as morphology is.

[zompist]

But then, this whole debate is sophistry. We know there is (obviously) a subjunctive mood in English. Saying "oh, but specifically in the past tense, for most verbs, this mood isn't morphologically marked, therefore for specifically the past tense it doesn't exist" is on a level with "[h] is an allophone of /N/". Or, indeed, the equivalent "Latin didn't have a present passive imperative, it just used the infinitive", or "Latin didn't agree with person, mood or tense in the passive perfect, except for the verb 'esse'". You CAN say these things, it just has no useful meaning or purpose. It's playing with words.

You know, there are other people in the world besides yourself, and sometimes they have other viewpoints. So please don't dress up your opinion as "we know (obviously)". If there is a debate you do not get to win it by declaring that your point is obviously true and any other is "sophistry".

In morphology, it is not uncommon for just a handful of words to exhibit behavior that other words of that class do not. It is by no means "obvious" whether to treat these as anomalies, or as establishing a general class.

I'll give an example: does French have two genders or three? There are two obvious frames:

"Le X est bon. Les X sont bons."
"La X est bonne. Les X sont bonnes."

But there's a third frame as well:

"Le X est bon. Les X sont bonnes."

But only three words fit in this category. In this case people don't really like to posit a third gender, so we just talk about exceptions.

What if this class had hundreds of members? Well, in that case we probably would speak of a third gender— that's what we do for Romanian, where singular/plural endings have three possible frames.

Or take gender in Russian: are there six genders or three? There are definitely six different agreement behaviors, but as they affect only one case it's easier to deal with this with a special rule (informally, "animates use the genitive as the accusative").

Or, are there genders on Russian adjectives in the plural? In fact the gender distinction is neutralized. You could certainly create tables of adjective inflection which include masculine/feminine/neuter forms for the adjective plurals, but it would be a waste of space and would create unnecessary confusion.

Similar situations come up with other categories of inflection, of course.

For deciding what categories of inflection exist, it's fine for some to follow tradition, or God's voice. But tradition is a mixed beast that is usually parochial, often ahistorical, and often just plain wrong. So we should always be willing to question it.

Most English verbs have just five morphological forms (six if you count 'to X'). Why complicate students' lives by insisting that they have another form which just happens to be identical to another form for EVERY SINGLE VERB except one? That one verb 'be' is just weird, and moreover it's entirely normal and common for 'be' to be weird.

With synthetic forms, like 'I have gone', we should remember that we have a mostly-synthetic language: there is no need to create a long list of paradigmatic forms as if we were speaking French— it suffices to give the rule for regular constructions.

(To answer cntrational, who posted while I was writing all this, of course you can have all sorts of syntactic constructions, and it may even be useful to label some of them 'subjunctive'. But there are a lot of constructions in English, and there's really nothing special about those that happen to be lexical forms in Latin. E.g. in American English we can say "I was captured" or "I got captured". That is, we have at least two passive constructions. It's still not a reason to up the count of morphological forms.)