[User John Lawler]: So how many named special cases of each modal verb are there, in toto? There's
1. subjunctive should, and
2. mandative should, and
3. putative should, at least;
and I s[up]pose 4. Should we? is suggestive should, and
5. Shouldn't we? is tag **should**, and I can think of a lot more.

I ask about only operative uses of this modal verb including literary and formal usage, but please omit anything obsolete. I'm trying to classify them and learn the formal terms describing each.
Are there any apt others in addition to the ones above? (I added the enumeration below)

1. What's 'subjunctive should'? Whenever should is conjugated in the subjunctive mood?
5. What's a tag **should** ?
6. Where does should/would feature? What's it called? This should is detailed in point 3 under 'Usage Notes' on page 1 of 2 or the last para herein or "should like to" vs. "would like to".

  • 4
    The conclusion of John Lawler's original comment seems to be that pursuing these distinctions offers very little value.
    – ScotM
    Jan 27, 2015 at 23:18
  • 1
    I think I would agree with that. Collecting "meanings" for modal auxiliaries is a losing game. Classifying them doesn't tell you how to recognize or describe them. As the OED and Cerberus demonstrate. Jan 27, 2015 at 23:38
  • 2
    So your trouble is that should is used in diverse ways, and this confuses you? This really is true of all modals. My recommendation is that you try to learn them the way a native speaker learns them instead of trying to tally and classify them all.
    – tchrist
    Jan 28, 2015 at 2:52
  • 3
    @LawArea51Proposal-Commit Read a lot. Listen to people talking. Eventually it all kinda falls into place and you don’t even notice that it is being used in so many different ways, because your mind will come to encompass the polysemic superposition without thinking anything of it.
    – tchrist
    Jan 28, 2015 at 3:19
  • 3
    @tchrist My goal for today will be to nonchalantly drop "polysemic superposition" into conversation.
    – Dan Bron
    Jan 28, 2015 at 15:52

2 Answers 2


The Oxford English Dictionary, under shall, describes 11 different special uses of should (#12–23), some of which obsolete, but most having several sub-uses. And this excludes many of the regular uses of should as the ordinary past tense of shall, some of which are included in #24–29.

The past tense should with temporal function.

  1. †Expressing a former obligation or necessity: = ‘was bound to’, ‘had to’. Obs.

  2. †In statements of what was formerly intended or settled to take place; = ‘was to’, or (contextually) ‘was about to’. Obs.

  3. Used in indirect reported utterances, or other statements relating to past time, where shall would be used if the time referred to were present.

    a. corresponding to shall in sense 5, 6, or 7.

    b. corresponding to shall in sense 8.

    Here should is the auxiliary of the ‘anterior future’ or ‘future in the past’ tense. With perf. inf. it forms the ‘anterior future perfect’ or ‘future perfect in the past’.

    c. in hypothetical, temporal, and final clauses, and relative clauses with hypothetical or final implication. (Cf. 10.)

    d. In noun-clause dependent on expressions of willing, desiring, commanding, requesting, etc. (in the pa. tense). Similarly (esp. with the verb want) in the pres. tense (colloq., orig. and chiefly U.S. and in representations of Jewish speech). (Cf. 11 and 22a.)

    e. In statements of a former likelihood, unlikelihood, expectation, hope, fear, etc.

    In present usage the rules for the choice of the auxiliary are the same as apply to the future tense (see 8). Until the middle of the 19th c., however, should was common in this use in the second and third persons, where would is now normal.

    f. In statements of what habitually occurred. (Cf. sense 9.) Now rare (? dial.).

  4. a. Forming with the inf. a substitute for the pa. tense ind. (or, with perf. inf., for the pluperf.) in the oblique report of another's statement in order to imply that the speaker does not commit himself to the truth of the alleged fact. (The perf. inf. was often substituted for the pres. inf. merely in order to express the notion of past time more unambiguously. Obs. exc. dial.

    The corresponding use of shall (= G. soll, ‘is said to’) is not evidenced in Eng., the OE. instances alleged by Bosw.–Toller having apparently a different meaning.

    ¶b. with omission of the have of the perf. inf.

  5. In indirect question relating to a past matter of fact. Obs. exc. arch.

    Present usage prefers the pa. tense or perf.; when the notion of uncertainty is emphasized, might or could is used instead of the earlier should.

  6. In questions introduced by who, whom, what, and followed by but, serving to express the unexpectedness of some past occurrence.

The past tense should with modal function.

As with other auxiliaries, the pa. tense (orig. subjunctive) of shall is often used to express, not a reference to past time, but a modal qualification of the notion expressed by the present tense. Where in addition the notion of past time is to be expressed, this can often be effected by the use of the perf. instead of the pres. inf. (though sometimes this produces ambiguity); the temporal notion may however be merely contextually implied, and in that case the pa. tense has the appearance of having both functions (temporal and modal) at once.

  1. a. In statements of duty, obligation, or propriety (originally, as applicable to hypothetical conditions not regarded as real). Also, in statements of expectation, likelihood, prediction, etc.

    This conditional form of expression was from an early period substituted for the unconditional shall in sense 2, and in mod.Eng. the pres. tense in this use is obs., and should = ought to.

    ¶with omission of have in perf. inf.

    b. SHOULD BE: ought according to appearances to be, presumably is. Also, ought according to expectation to be, presumably will be (cf. sense 18a).

    c. YOU SHOULD HEAR, SEE = I wish you could hear, if only you could hear, etc.

    d. Used ironically, expressing the inappropriateness or unlikeliness of the action advocated or state envisaged, as I SHOULD WORRY, there is no reason for me to worry, I am not worried. colloq. (orig. a Yiddishism).

  2. In the apodosis of a hypothetical proposition (expressed or implied), indicating that the supposition, and therefore its consequence, is unreal.

    a. Where shall (in sense 5, 6, 7, 8, or 9) would be used if the hypothesis were accepted.

    b. When the pres. tense of the principal vb. would be used if the hypothesis were accepted. (Where the pa. tense or the perf. would be used, should is followed by the perf. inf.)

    In this use the combination of should with inf. forms a periphrastic past subjunctive: thus ‘I should be’ = the archaic ‘I were’. Similarly with perf. inf.: ‘Then I should have been’ = ‘then had I been’.

    The choice between should and would follows the same rules as that between shall and will as future auxiliaries, except that should must sometimes be avoided on account of liability to be misinterpreted as = ‘ought to’ (sense 18). In present Eng. should occurs mainly in the first person; in the other persons it follows the rule for shall in 8 c, d.

    ¶with omission of have in perf. inf.

    c. With verbs of liking, preference, etc., should in the first person (and interrogatively in the second) is regarded as more correct than would, though this is often used.

    In the third person should is used only in indirect speech (when he represents I); uses like quot. 1862 are abnormal.

    The forms I should have liked to (see) and I should like to have (seen) are alternative ways of adding the temporal notion to the modal sense of should. Another form, sometimes met with, but certainly faulty, is I should have liked to have (seen).

    d. The original conditional notion is obscured in the phrases IT SHOULD SEEM (see SEEM v. 7f); one should think (now somewhat arch. and perh. sometimes interpreted in the sense of 18). Similarly I SHOULD THINK (SUPPOSE, etc.) = ‘I am inclined to think (suppose, etc.)’; also colloq. as a strong affirmation in reply to a tentative suggestion, e.g. ‘I should (rather) think he did object’.

    In the last phrase (as used idiomatically), would is never substituted; in the second person the phrase is used only in questions, and in the third person only in oblique narration.

    †e. SHOULD HAVE BEEN = ‘would have had to be’: see 3b. (In quot. with omission of have.)

    f. I SHOULD (DO so and so): orig. with expressed or understood protasis ‘if I were you’, but in mod. colloquial language often used loosely = ‘I would advise you to (do, etc.)’.

  3. In a hypothetical clause expressing a rejected supposition.

    †a. Where should has notional force = ‘were obliged to’, ‘must’, ‘were about to’. Often with ellipsis of if after as. Obs.

    With the use as in quot. 1530 cf. the modern ‘as if his heart would break’.

    b. Where the future tense (or the present with future import) would be used if the supposition were entertained. (With pa. tense subjunctive, usually should or would, also could, might, arch. were, etc., in the apodosis. Cf. 21.) Now somewhat rare, mod. usage preferring were to.

    †c. With reference to the past (e.g. ‘if he should have done’ = if he had done). Obs.

    d. In relative clause with hypothetical import.

    e. as WHO SHOULD SAY [cf. F. comme qui dirait] = as much as to say. arch. Also †AS IF WE WOULD SAY (SHOULD HAVE SAID).

  4. a. In a hypothetical clause relating to the future, should takes the place of shall (indicative or subjunctive), or of the equivalent use of the present tense, when the supposition, though entertained as possible, is viewed as less likely or less welcome than some alternative. (With future, future perf., or imperative in the apodosis.)

    b. Similarly, with perf. inf., in a hypothetical clause relating to what may have happened in the past.

  5. In a noun-clause (normally introduced by that).

    a. In dependence on expressions of will, desire, command, advice, request.

    Where the verb of the governing clause is in the pa. tense, this use is indistinguishable from that treated in 14d.

    The substitution of should for the earlier shall (itself a periphrastic substitute for the more primitive use of the pres. subjunctive: see 11a) may have arisen from instances in which the governing vb. was in the modal pa. tense (as in quots. c1200, 1340).

    b. In statements relating to the necessity, justice, propriety, etc. of something contemplated as future, or as an abstract supposition.

    c. In expressions of surprise or its absence, approval or disapproval, of some present or past fact.

    ¶with omission of have in the perf. inf.

    d. In clause dependent on sentence (negative, interrogative, or hypothetical) expressing possibility, probability, or expectation.

    Cf. ‘Is it possible that he should do this?’ with ‘It is possible that he may do this’. Similarly, ‘It is unlikely that he should have been there’, but ‘It is likely that he was (or may have been) there’.

    e. In clause (now almost always with lest) expressing the object of fear or precaution.

  6. In special interrogative uses.

    a. In questions introduced by why (or equivalent word), implying the speaker's inability to conceive any reason or justification for something actual or contemplated, or any ground for believing something to be fact.

    b. In questions introduced by how, implying that the speaker regards something as impossible or inadmissible.

    †c. In questions relating to meaning, cause, or reason, the form with should was formerly often substituted for an indicative tense. Obs.

Elliptical and quasi-elliptical uses

  1. With ellipsis of verb of motion: = ‘shall go’. Now arch.

    The use is common in OHG. and OS., and in later HG., LG., and Du. In the mod. Scandinavian langs. it is also common, and instances occur in MSw.

  2. †In questions, WHAT SHALL = ‘what shall (it) profit’, ‘what good shall (I) do’. Obs. (rare after OE.).

  3. †With the sense ‘is due’, ‘is proper’, ‘is to be given or applied’. Obs. Cf. G. soll.

  4. a. With ellipsis of active infinitive to be supplied from the context.

  5. †a. With ellipsis of do (not occurring in the context). Obs. rare.

    The place of the inf. is sometimes supplied by that or so placed at the beginning of the sentence.

    The construction may be regarded as an ellipsis of do. It is distinct from the use (belonging to 27) in which so has the sense of ‘thus’, ‘likewise’, or ‘also’; in the latter there is usually inversion, as so shall I.

  6. †With ellipsis of be or passive inf., or with so in place of this (where the preceding context has is, was, etc.). Obs.

Image of OED online from Chrome on Windows7 showing formatting (can be removed). That's what I see for shall 20.e (my yellow highlighting) enter image description here

  • Why the small-caps? OED don't do small-caps for those uses they only use it for names of authors (and links to other words).
    – Frank
    Jan 28, 2015 at 15:56
  • @Frank Actually, yes they do. Even if you have no subscription, go to the OED’s online site and look at the word of the day and the recent entries under it, all of which require no subscription. You'll see them using small caps for exactly these things that are in the posting.
    – tchrist
    Jan 29, 2015 at 22:26
  • @tchrist Cerberus: Thank you for your comment. I wrongly used 'mock' in an old post and instead had intended to mean whether the earlier comments meant to imply that my question was unhelpful. I admit that I posed a bad question so have tried to recast it, but I shall peruse this. I apologise for any offense.
    – user50720
    Jan 30, 2015 at 3:59
  • @tchrist - The online OED (that I see) uses bold-italic normal cased letters where you've used small-caps. Maybe it's a browser difference, just wondered. I've added an image of what it looks like on Windows7/Chrome - feel free to remove the image any time. This should probably be in meta really.
    – Frank
    Jan 30, 2015 at 6:12
  1. Subjunctive meaning of should:

Should you get lost in Mexico, remember el baño means bathroom.

  1. Mandative (AKA Obligatory)meaning of should:

You should eat your vegetables!

  1. Putative meaning of should:

It struck us as funny that you should ask this question.

  1. Suggestive meaning of should:

The others brought a flashlight along. Should we?

  1. Tag meaning of should:

He shouldn't have another drink. Should he?

These subtle modal auxiliary meanings of should carry on ad nauseam!

  1. Both should and would have been doing so much modal auxiliary work for so long in English that we often find them working together and doing each others jobs.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.