Yesterday, I read about the English subjunctive mood. I tried to, but couldn't, discern a concise conception of it.

What do you regard as a useful and concise conception of the subjunctive mood?

  • 4
    A useful and concise conception of the English subjunctive mood is that it is a mythical beast, like the fairies at the bottom of the garden in summer. English has no subjunctive mood, though Latin did and many European languages (Spanish, German, French, etc) do have a subjunctive mood. But English teachers talk about it all the time, and often have faith that it exists; just like fairies. In fact, though, as you've discovered, there is no simple description; there is only a name and a lot of vague hand-waving about "conditional" and "hypothetical", like they were detectable. Mar 6, 2014 at 17:51
  • Monsieur Lawler, I shall debunk your stmt tonite, thro the lens of Mathematics (after all linguistics is a mathematical way of studying languages). Mar 6, 2014 at 18:00
  • 3
    There's no concise conception of it today, because today there are two different remnants of the subjunctive in use: the mandative subjunctive as in "and a thousand faces swirled around me, clamoring that my life be burned" and the irrealis or were subjunctive: *He knew the names of birds, and mocked their notes, and whistled, as if he were a bird himself. * Mar 6, 2014 at 18:04
  • 1
    @JohnLawler the appendix is a vestigial organ, but it's an organ nonetheless … I'm not disagreeing wholesale, but there are remnants of a once proud subjunctive. I will agree, though that most of them are clumsy approximations of the usage in romance and other languages.
    – David M
    Mar 6, 2014 at 18:13
  • 1
    And as an appendum to my previous comment, the subjunctive lives on in a few idiomatic phrases, such as "long live the King". But without the earlier diverse uses of the subjunctive that connected them, I think you might as well consider the mandative subjunctive and the irrealis as two different things. Mar 6, 2014 at 20:48

5 Answers 5


Regardless if a language supports it, the subjunctive exists.

The ambiguous, vague and imprecise view of phrase-structure linguistics (vs the more precise dependency structure linguistics) calls the subjunctive a "mood". Unfortunately, prevalent English grammatical analysis is still driven by the obsolete phrase-structure grammar. Therefore, we would find the subjunctive concept ill-defined under this outmoded grammatical analysis framework.

In Mathematics, we have a concept known as imaginary numbers. Imaginary numbers are just as concrete in existence and equally useful as "real" numbers. Where, square root of "−1" is the base factor of imaginary numbers.

The subjunctive exists in imaginary time. Using computer science lingo, the subjunctive is an encapsulated object.

What is an object?

Let's look at the Seawolf class of submarines. Seawolf class is just a design. It may stay just as a design without having any submarines built. So that we could merrily refer to the Seawolf without any actual submarines built. But when we refer to the USS-Connecticut, we would be referring to an actual object built to the specs of the Seawolf design.

What is a module?

When we design integrated circuits, we have blocks of design templates stored in our library of structures. Each template encompasses not just the circuitry but also constraints of their administrative and manufacturing processes. And then we could start picking those templates from the toolbar and lay them out to arrange into an actual IC design.

Time capsule grammar?

What if we could create a descriptive template of events, where all the past, present, future are related to each other in an independent encapsulation box, as though time other than within that box does not exist? And then we could use this template and drop it into any actual time past, present or future.

Such a grammatical template is what we would call the subjunctive "mood". Notice that I condescendingly and subjunctively refer to "mood", in disdain of its vagueness. A subjunctive encapsulation would have the full set of past, present, future tenses, progressive, perfected, infinite elements, participles, etc.

Subjunctive in English:

However, in English, we do not have a separate set of tenses for the subjunctive encapsulation. We use the past to denote subjunctive encapsulation. Using the past does make some sense, as it fulfills the functionality of yanking a module off from being grounded to any actual time, to allow us to freely deploy the module at any point in time.

Why do we need temporally (aka time-based) ungrounded modular grammar? Because we want to be able to move a set of encapsulated events freely across time, while keeping the relativity within the encapsulation intact:

  1. We want to talk about fantasies, which we know will never happen at any time.

  2. We want to talk about actual events that happen all the time, but we just wish to talk about one exemplary instance without restricting when in history they would happen — regardless now, in the past or in the future.

  3. We want to talk about actual events that either happened in the past, are happening now, or will happen in the future, but which happen regularly across a restricted range of time. But, we wish to describe an exemplary point in time without restricting at which actual point in time.

  4. We want to talk about actual events that either have happened or will happen, but we wish to hypothetically shift those events to another point in time.

  5. We want to propose bringing into existence events which are possible, but which have not happened before.

  6. We want to propose bringing into existence events which have happened to someone, some other place, in a story or in a movie — specifying where and when we wish to have them happen.

  7. We want to bring into existence a set of events A, if and only if another set of events B happens. We don't know when set B would happen, or if would ever happen. And we wish to say that set A's happening is dependent on set B's happening, stating either, A before B, A with B, A after B, or A just happens at an indefinite time as long as B happens.

  8. etc, etc.

So I gave 8 examples of why we need the subjunctive time module and the most important example is example #8. Why?

You see, the need for encapsulated time module is a continuum. All the space on the planet would not be adequate to document all the reasons and situations you would need to use it. Traditional grammar would try to quantize that continuum into buckets like optative, cohortative, conditional, jussive, blah, blah — where frequently we would find a situation that would be inadequately contained by these quantization, or that a situation is either a composite of or in between of such quantization.

It is said that Taichi, when mastered, is a very potent and deadly set of skills. People describe it as unrestricted in time and space, or fluent as water. They compare it to Shaolin which is also said to be potent and deadly, but restricted and grounded in time and space. Imagine you could modularise your actions and responses, to fluently shift to deploy each module in time and space. So, people hyperbolically say when you master Taichi that way, you would become invincible.

So, similarly I urge y'all not to restrict your understanding of subjunctive modularity to those "moods" pigeon-holed by traditional constituency grammar. Free your mind, and just think about floating your subjunctive description in terms of either

  1. You wish to continue your event module ungrounded.
  2. You wish to ground the module to actual time, but without defining when.
  3. You wish to ground your module at a particular point in time.

Subjunctive language is very important in statistics, especially when Bayesian statistics is involved. Analysing the subjunctivity and event-relativity of each statistic is important. Whether a statistic describes events that happened, will happen, might happen, conditionally happens, etc. Unfortunately, most statisticians simply use the future tense, at least those whom I have interacted with.

Having freed your mind, you could then indulge in subjunctive Kabbalah:

  • A subjunctive of another subjunctive.
  • A subjunctive of another subjunctive, which is a subjunctive of another subjunctive. A cascade of subjunctives.
  • Bayesian dependency of subjunctives.
  • Cyclically dependent set of subjunctives.
  • Subjunctive recursion.
  • and most importantly of all: etc, etc.

The length thus far of this thesis gives me no space to give examples quantized to traditional constituency grammar — sorry. It is unfortunate that the English language does not have a separate set of tenses for the subjunctive. And therefore we have to skillfully use existing tenses and common-sense to describe subjunctive situations, on a case by case basis. I am certain there isn't a natural language on the planet that provides adequate means to treating subjunctive situations.

~ May the subjunctive modular karma be with you.


In response to the originally posed question, John Lawler wrote a somewhat amusing, (pedantically) accurate, but highly confusing comment, one which I feel deserves translation and further elaboration into something more resembling an actual answer. John said:

A useful and concise conception of the English subjunctive mood is that it is a mythical beast, like the fairies at the bottom of the garden in summer. English has no subjunctive mood, though Latin did and many European languages (Spanish, German, French, etc) do have a subjunctive mood. But English teachers talk about it all the time, and often have faith that it exists; just like fairies. In fact, though, as you’ve discovered, there is no simple description; there is only a name and a lot of vague hand-waving about “conditional” and “hypothetical”, like they were detectable.

What I believe John meant (and he is free to correct me) is that English no longer uses a special inflection for subjunctive situations which viewed in perfect isolation is clearly distinct from a bare unmarked infinitive, or uniquely with were-hypotheticals from an otherwise normal (albeit plural-looking) indicative inflection.

This stands in contrast to the situation in Latin and its descendants, and often in other Indo-European languages as well, where one can always distinguish an indicative from a subjunctive because of those respective tongues’ rich inflectional morphologies. For example:

  • Latin example: Starting with amare, compare 3sg present indicative amat with 3sg present subjunctive amet.
  • Spanish example: Starting with amar, compare 3sg present indicative ama with 3sg present subjunctive subjunctive ame.
  • French example: We can’t here use the obvious cognate aimer for our demo, because French no longer has a 3sg present ind/subj distinction there. Instead we will start with venir, where we can compare a 3sg present indicative vient with 3sg present subjunctive vienne.

The French aimer (and indeed, most regular French verbs) presents us with an interesting situation: a special subjunctive inflection is seen only in the 1st and 2nd person plurals (for which it “borrows” the corresponding imperfect indicative form, rather quite oddly enough), not in singulars or in the 3pl.

Does that mean that French no longer has a subjunctive for verbs like aimer where all forms of the French present subjunctive look just like either present indicatives or else just like imperfects. In John’s framework, I suspect that it indeed may not. I also suspect he’d never get a native French speaker to agree with him on this point.

I bring up the French case because there are some parallels with what has happened with English. But while French has lost a few of its distinctions and inflections, English has lost almost all of its original inflectional morphology all across the board, including the entire subjunctive set sauf for one unique situation alone.

So in the restricted sense that other languages have these specially inflected verb forms for the subjunctive that look nothing like their corresponding indicatives, no, English does not have that. John is 100% correct in this. But it is in a specific domain and not the one that most people coming here looking for answers are used to operating in.

Nonetheless, English has maintained a distinction between a normal indicative and an unmarked form in the 3sg case (the only place we ever make a present indicative inflection change), a distinction it makes for many of the same situations that in other languages have a mandatory inflection change.

This is easily demonstrated with the minimal semantic pair:

  • I insist that she is here.
    (Meaning: she is really here, and I am affirming that)
  • I insist that she be here.
    (Meaning: she is not here, and I am demanding her attendance, or to subjunctivize it, I am demanding that she attend)

So English does, in certain particular situations, make a distinction between the form used for subordinate clauses in a simple indicative statement of fact on the one hand, and the form used in counter-factual situations on the other. This is what people are talking about when they talk about “the subjunctive” in English. But it is a wholly invariant distinction that looks exactly like the unmarked infinitive such as one would use with a modal auxiliary. And that’s why John says it is not a subjunctive.

Note how this is more change than French experiences for the majority of its 3sg cases in a clause requiring the subjunctive — after all, they use the very same word as in the indicative while we do not — and they would never stop calling it a subjunctive construction.

The odd man out in all this is were for hypotheticals. This is where the other languages I’ve mentioned (sometimes) use a special inflection that there they refer to as something like a “past subjunctive”. For example, those are “l’imparfait du subjonctif” in French (now almost exclusively a “literary” inflection, not a spoken one) and “el (pretérito) imperfecto del subjuntivo” in Spanish.

The thing is, in most of these other languages, they automatically backshift a present subjunctive to a past one when relating past events. But in English we never do that. That’s not what our were form is for. This special form is not a simple shift from present to past, since in English going from the main clause being in the present indicative

  • I demand that she be here.

to being in the past indicative

  • I demanded that she be here.

sees no change in the subordinate clause’s verb, only in the main one’s. In the other languages, it doesn’t typically work that way. So be and were do not oppose each other in the way many other languages’ subjunctive forms typically do.

Only in the most archaically stilted of examples, or ancient ones, can one ever find a tense shift from present to past in an English subjunctive clause. And even there it may be a learnèd import from other languages. English has never had a tense shift in the subjunctive in quite the same way as Latin and its descendants do.

However, we in English do use — as a unique and isolated case — the special form were for hypotheticals. This is another place besides back-shifting where neighboring languages also use a past subjunctive: they used it for hypotheticals.

So when people speak of a past subjunctive in English, they are talking solely about an “If only it were so!” or an “If I were you” type of hypothetical, never a back-shifted present subjunctive.

It’s easy to see why some people continue to call this a “past subjunctive” in English: that’s what the Old English past subjunctive form indeed was:

OE *to be* inflections

Or at least, that’s where our use of were comes from. Notice the past subjunctive wǣre for singular and wǣren for plural. If one concedes that Old English wǣre turned into Modern English were, and I rather think one must, then it becomes clear that we use were now just as we ever did. It is the same verb, and the same inflection, and for the most part, the same rules, as we have always used for these situations.

Considering how were inarguably started out as a past subjunctive back in its wǣre days, it should be no surprise that some people still call it that same thing today — even if that term may seem misleading to certain other folks.

It really all depends on where you’re coming from.

Additionally, if you look closely at the Old English inflection table above, you can all see where we came from for using be as a present subjunctive, and in the plural it was indeed identical to the infinitive. The situations where we use special forms haven’t changed as much as people would have one believe.

It is just this sort of diachronic analysis that can lead a person to speak of “English subjunctive forms”: the forms we still (sometimes) use today are historically linked directly back to the original, richer inflectional morphology in the Old English.

However, when examined under the light of synchronic analysis (which I presume is what John is doing here), the subjunctive picture blurs into something unrecognizable at best, and arguably even disappears altogether. Children of English-speaking parents do not learn a table of special inflections of subjunctive forms the way children of Romance speakers eventually do. That’s because English has no such thing.

So even though English once upon a time had special subjunctive inflections for certain situations (although it no longer does), and just because some speakers and writers maintain some of those same distinctions today unconsciously, it is imperative that one never attempt to mindlessly apply Latinate rules of where to use a present or past subjunctive into any form of English, not even in Old English.

That’s because English has never precisely followed those same rules. If you take some time to read through the copious textual examples regarding all this in the second volume of F.T. Visser’s monumental An Historical Syntax of the English Language, where the author gives OE, MidE, and ModE examples of what he calls a “modally marked form”, you will see that even a thousand years ago English marched to the beat of its own drum, not Rome’s, and that there was even then variation in writers’ choices.

I’ve answered a lot of subjunctive questions here. I might also recommend these other answers of mine for further, somewhat lighter reading:

  • Excellent answer, and thank you especially for explaining what Prof. Lawler means when he talks about the subjunctive! Mar 9, 2014 at 21:14
  • 5
    I’m not sure I really buy the argument that there is no subjunctive just because there is no separate table of inflections. The Scandinavian languages are usually said to have a subjunctive mood (though even more vestigial than its English counterpart), and they have no inflection for any person in any tense or mood. The present subjunctive is simply the infinitive, as in English. That doesn’t mean it’s not a subjunctive—it’s used only in archetypical subjunctival contexts, being invalid everywhere else. Latin doesn’t ‘own’ the morphosemantic category of subjunctivity, after all. Mar 12, 2014 at 0:16
  • 2
    @JanusBahsJacquet I’m in your camp on this one, not in John’s. I was simply translating his professorial ipse dixit into something that others might stand a chance at understanding, since at face value, to say that “English has no subjunctive” seems counter to their training and understanding. People have to understand that John is talking in John-talk, which is all about inflections here.
    – tchrist
    Mar 12, 2014 at 0:28
  • 1
    Oh, I realised that—my comment was, I guess, just as much a comment to him as a comment to your excellent answer. I never even thought about the fact that there is no subjunctive backshift in English before (there isn't in modern French, either, but that's for a different reason of course). Mar 12, 2014 at 0:50

The subjunctive mood expresses a hypothetical or counterfactual condition. It is perhaps best understood by looking at the difference between if I were (past subjunctive) and if I was (past indicative).

If I were a carpenter and you were a lady, would you marry me anyway?

The speaker is not a carpenter, but is posing a hypothetical scenario that supposes that he is a carpenter, and is asking a question based on that hypothetical scenario. (Note that was is often used in place of were these days, and with increasing acceptance from linguists. However, this does not change the fundamental nature of the subjunctive.)

If I was rude to you yesterday, I apologize.

This is also a conditional, but it is not based on a hypothetical scenario: either the speaker was rude in the past, in which case he apologizes, or he was not—in which case there is no need for an apology. (Could you use "If I was a carpenter" as past indicative? Possibly—if you've lost your memory and do not remember whether you did or did not work as a carpenter in the past.)

Understanding the difference between the two will help you explore the subjunctive further.

  • This is half of the subjunctive; the other half is in forms like “Let there be light.” Mar 9, 2014 at 9:10
  • 1
    @Bradd: "Let there be light" is subjunctive? Why? Isn't it the same construction as "Make him go home", which I don't think anybody would call subjunctive? Mar 9, 2014 at 14:48
  • Oops, did I get the mandative subjunctive wrong? Mar 9, 2014 at 20:59
  • 2
    @BraddSzonye: That was an infinitive, but you meant he decreed that there be light, which is indeed a present subjunctive. Mar 10, 2014 at 2:06
  • Isn't it still a hypothetical with, If I was? I would use, If I were rude/should should. Because this is not an apology to say "If" first. It is a werepology.
    – sas08
    Apr 8, 2019 at 22:40

The English subjunctive still exists today, and it is still very rife and proud; however, because English verbs have syncretized over the centuries, the entire subjunctive conjugation has fallen off and merged with the conjugations of the indicative mood paradigm of the verbs with the exception of a few vestigial traces--i.e. the "to be" verb for the entire paradigm in the past and present subjunctive, third person singular forms, and negations, and tense agreements, e.g. (I demanded that he do it.).

One of the reasons for the syncretism of the English language is due in large part to the nine modal verbs we have today: shall, should, will, would, can, could, may, might, must. These modals have replaced much of the subjunctive mood's usage in the language; where it once was, these rogues have filled the empty spaces. Other languages like Latin and French don't have modals; they use the subjunctive quite frequently where we would use our modals. It's just a fact of the matter. If it weren't for the modals, English would be lost as a language. If you think about it, we don't have a future tense of a verb. In order to create the future tense, we have to take the present indicative form and place "shall" or "will" or some other periphrastic construction in front of the main verb to make the sentence into the future. In English, we say, "I shall buy the book"; in French, "J'acheterai le livre". French has a future conjugation for its verbs; English does not; thus we need the modals. What would we do without them?

So if you should ever wonder again about what has happened to the subjunctive in English, just remember that you say it every day, perhaps unknowingly, just by using many of those modals above. In fact, the word "modal" comes from the Latin word for "mood". It's important that we not spend too much time quibbling on such small issues and points. Whether you use the subjunctive or not, your speech probably won't become unintelligible; although there are times when failure to use the subjunctive can change the entire meaning of one's sentence to something he doesn't want it to be. We shall not go into that right now as it is unimportant. I hope this was helpful for you.

  • Both Latin and French most certainly do have modals—they just don’t use all of them in the exact same way as English. ‘I should go home’ and ‘je dois rentrer’ both use the same modeme (if that’s a word) with an infinitive complement, for example. Mar 12, 2014 at 0:21

Okay, to Janus Bahs Jacquet, when I said that Latin and French don't have modals, I meant that they don't have "modals" in the same use as English. You give "dois" as a conjugated form of "devoir" meaning "should" in English; however, "devoir" acts more like an auxiliary verb rather than a modal auxiliary, kind of like "do" or "be" or "have" in English. We don't conjugate modals in English except in 2nd person archaic singular and I guess one could argue in the past tense forms (past tense of "shall" is "should", etc.). English has strict modals to replace the subjunctive forms and other indicative forms such as the future tense, which doesn't have a verbal conjugation in English such as in French "devrai", which is roughly translated as, "I shall have to".

It's just the way it works, Janus. I am not saying that you are wrong and I may have misstated the use of "modals" per se, but I would still argue that modals in Latin and French may be closer to "auxiliary verbs" rather than "pure" modals; however I do concede that it is possible that I might be wrong herein. I did major in English and have taken both Latin and French and have never heard "devoir" or any other auxiliary referred to as a "modal auxiliary"; however that doesn't mean that they aren't referred to as such and I do agree that "devoir" is translated into the English verb "should" in many situations wherein "should" is used, but not in every situation.

Since this article is about the "subjunctive mood", we can both agree that "should" replaces the present subjunctive in English and is actually the past subjunctive conjugation of "shall"; therefore, in a sentence like, "If I should die...", we can both agree that the "literal" translation of, "Si je devrais mourir...", would mean something completely different from its English equivalent above; well, at least, I should hope you would agree. If not, you may voice your opinion herein. Personally, I believe the proper translation of, "If I should die...", into French would be, "Si je meurs...", but I don't speak French fluently and didn't major or minor in it; I have just taken many classes of it over the years of my studies. If you speak French fluently, please comment on my translation.

Oh, and I couldn't find "modeme" as a word online or in any dictionary so I doubt it might be a word. You should ask Oxford to adopt it, though; it appears to be useful.

Also, to Bradd Szonye, "Let there be light" and "make him go away" are both subjunctive. This is called the "hortatory subjunctive" in English. We can extract both "be" and "go" therein by rewording the statements to "I command that there be light" and "make it so that he go away". There you go. Don't listen to the other two who tell you it's just a bare infinitive; they're just as confused as many grammarians are on this subject.

  • French is not a very good example of using things that take a past subjunctive. It has lost that. Contast Si me muriera, no tendría nada que decir — If I were to die, I’d have nothing to say.
    – tchrist
    Mar 13, 2014 at 22:55

Not the answer you're looking for? Browse other questions tagged or ask your own question.