REPLY of 7/3/96 TO E-MAIL INQUIRY received on same date

ON PRONOUN CASE FORMS AND 
VERB MODALITIES IN ENGLISH

[updated 5-1-99]

© 1996, 1999 by Orchid Land Publications

      Thanks for writing.

      Some of the work written up will need to be translated by someone engaged at the appropriate grade level. My Middigrammar has been used in various places, but is still far from its final form and hence is not available. I can refer you to the booklet, Why more English instruction won't mean better grammar listed on the Orchid Land Publications website; it will perhaps prove more useful in a school situation than Variation in the data--which contains more information in some of its chapters. You may even wish to read chapter six of my forthcoming Oxford University Press book--which should be out by now but apparently isn't yet. (They're charging $98.00, but it is likely to be tax-deductible if you are a teacher.)

     Whether you happen to be a highschool teacher or graduate student, the usefulness of this information depends on whether your personality and background allow you to shift gears from the static approach (and the "Let's make a list of usages" approach) of the typical grammars to an approach of "Let's make a system that explains the list"--a dynamic approach with dynamic modalities--those that alter the position or presence/absence of a form or the choice of a form according to the context or force of an utterance. I haven't got time for a long piece, but I will try to illustrate briefly what is involved. Before proceeding, let me take note of the reason why grammar mystifies ordinary people and teachers of English: There is a great chasm or disconnect between what grammars say and what educated speakers of English do; it's unreal.

     FIRST, why do people who say "for me" and "for her" nevertheless believe it is correct to say "for she and I"? Further, why do people who would never say "myself did it" say "John and myself did it"? Why indeed did a prime minister of the United Kingdom (and former university professor) say on the BBC, "to we Europeans," when he would never have said "to we" without the appositive? Similarly, the announcer who said (in an ironic or humorous vein) on the BBC Letter Box, "Us humans get very attached to our pets," was just saying what is normal English for many people; note the appositive. I hope you get the idea of the problem--before we proceed to deal with a solution. Many believe that "to her and me" sounds illiterate. To see whether they are right, check below. Many persons that one hears on media shows say "Her 'n me did it" in the Romance-language manner (cf. Adrian Mole's "My Father and me [did it]"). (On "Don and I's baby," see Variation in the data, p. 100). The casual style of "Him and me did it" would be mostly avoided in educated adult speech, where (equally Romance-language) "and myself" would often be heard.

     A recently heard example, "Let he who is without sin throw the first stone," may not be a clear-cut example. For writers in England and elsewhere were using such examples a century ago, possibly treating let rather like the optative use of may in "May they win"--which quite reverses the potentiality of "They may win." Since, on the other hand, one wouldn't say "Let he win," one must assume that a relative clause acts like an appositive, creating here a marked environment for the reversal of a personal pronoun's case form. (Cf. further unmarked "Paul, the king" with marked "King Paul," where "King" is a title; the order of the appositive and the use of "the" with it are reversed.)

     But what cultivated native-speaker would point to her picture in the newspaper and say, "That's I!"--as Marilyn Monroe was made to say in Gentlemen prefer blondes in order to look silly and speak the way foreigners are taught to use English--by the book or "standard"? But if "That's me" is correct English, as it is, then the 1611 Bible's "Whom say ye that I am?" is correct after all--though the complementizer that would be anomalous here in current English.

     SECOND, why do "He got shot"--with get representing a non-process (a report or state)--and "It was/is (being) aimed at"--with be representing a process--reverse these two aspectualities in the following?

--POSTERIOR:
He was/is being shot at dawn tomorrow. (non-process)
It was/is getting aimed at for a take-over tomorrow. (process)

--ANTERIOR (with have):
He has/had been shot. (non-process)
He has/had been getting (or just: gotten) talked about. (process)

--EXOCHRONOUS (timeless):
She is devasted. (non-process)
It gets talked about a lot. (process)

     Why can't we say "It has already been done at midnight" although the following is okay: "Every time I get there at noon, it has already been done at midnight"? A thoughtful reader will realize that the answer must jibe with the fact that we say to have done it at midnight (not *to was/were done it at midnight) and having done it at midnight (not *wasing/*werering done it at midnight). In other words, we are dealing with categories that are marked. The examples could be multiplied with not only passives. On your own, try contrasting "being caught" and "getting caught" with, respectively, "to be caught" and "to be getting caught" or "to get caught." (Imperatives are discussed below.) Compare, e.g., the force of "That gets talked about a lot" and "That can be talked about a lot"--where the modal verb, can, creates a marked verb phrase. (For the linguist: Doubly marked conditional could reverses the uses found with irrealis can; and deontic and epistemic forces exhibit reversed usages.)

     Participles are neutral, while infinitives are special--what below is called marked. Also marked are the three modalities, or verb categories, just discussed--posterior, anterior, and exochronous (timeless), as well as the "surrealis" environment illustrated below. (It is a subordinate clause--conditional, temporal, adversative, relative, but not causal--embedded in a clause with posterior [including imperative] predicate.)*

     I don't wish to "ironize" too much, but why do grammarians of English speak of a present-perfect "tense" despite these three facts:  (1) Languages having true perfects (forms for completed acts or states) and a progressive modality cannot combine the two, as they are not logically compatible.  (2a) Languages having a perfect form expressing finished acts or states cannot, for obvious reasons, say things like the English example, "It has stood [or has been standing] here for ten years and hasn't fallen down yet."  (2b) Statements like the foregoing English example are in the present tense in languages which lack semantic-pragmatic parallels to the English forms with have. (In the formal aspect of languages, things are changing in the other direction, since semantic pasts in many European languages are being formally replaced by compound or phrasal verbs formed on their auxiliary verbs for "have" and/or "be".) Note that, while have-formations are holding their own in English in marked categories like infinitives, participles, etc., it is the other way around in unmarked situations: For English have-formations are giving way to past formations in unmarked situations; cf. "They already did it" and "They didn't come yet". (Of course, "They just arrived" makes for better system logic than the version with have!)

     Why does should express the volitional or intentional force of doubt or uncertainty ([- expectative] = "dubitative") in

--If they should arrive on time
--Whoever should be the one selected
--Though I should walk through the valley of the shadow of death

but factual force in

They were surprised

Why does shall express a mild suggestion in:

Shall we leave now?"

rather than the volitional or intentional force of determination that is palpable in:

We SHALL overcome!

     Why do be gonna and will ('ll) switch forces when we negate them or question them (not in a rhetorical question), as in the following?

--UNMARKED:
It's gonna rain. (unmarked force in unmarked category)
It'll rain. (settled notion: marked force, fairly unmarked set of forms in marked posterior category)

--MARKED (negation):
It's never gonna rain. (settled notion: marked force in marked category)
It won't rain today. (unmarked force in marked cagtegory)

--MARKED (interrogation):
Are they really gonna be on time? (settled notion: marked force in marked category)
Will they be on time? (neutral unmarked force in marked category)

In unmarked (default) categories, will represents the marked force of a settled notion; but be gonna acquires this force and will becomes neutral in marked categories or situations. Of course, the be- and get- passives also switch the aspectualities that they signal when they are found in negative, true-interrogative [not rhetorical questions], and comparative [than- clause] environments--switching from what would be expected in the corresponding non-negative, non-interrogative, and non-comparative situations. Be gonna is the only posterior in marked categories like infinitives and participles--which, like modals, are exochronous or posterior by definition.

      Why do we consider rings unmarked in "The bells rings at ten" when we add "every day" (exochronous, i.e. timeless) but regard it as marked when it's posterior in "The bell rings at ten tomorrow" (a schedule or routine being indicated), whereas in a doubly marked posterior situation (the "surrealis" clauses in the following illustrations), rings is, again, unmarked (in a doubly marked environment):

When/As soon as/Till the bell rings tomorrow, . . .
If it rings at ten tomorrow, . . .
Give it to whoever rings the bell tomorrow

(In such examples, be gonna and will would be marked in having special force; e.g. "When the bell will ring, If the bell's gonna ring at ten," etc. This force is that of "expected to occur" or a settled-notion anticipation. (Because is different: "They'll be tired tomorrow because they will have hiked too far." Will is the same here as in other unmarked situations.) To create the opposite force, that of "not expected to occur," we throw the time [of the neutral expression, which is in the exochronous modality in the "surrealis" environment] back, as in "If she did it by tomorrow, . . . ; If he were to arrive on time tomorrow, . . .; If she'd arrived by then tomorrow, they would of course be pleased." (Historically, at least, "If she should arrive on time tomorrow, . . ." fell into the same ballpark. Compare the neutral expression: "If it arrives tomorrow, . . ." Note that, because of settled-notion force, will is appropriate in the main clause of a conditional; for the conclusion of the main clause follows from the validity of the subordinate if-clause.) The result of these variants is that in doubly marked posterior hypotheses clauses, two forms of dubitative ([- expectative]) hypotheses and one overmarked counterexpectative hypothesis clause are distinguished from the unmarked or default format. Using will in surrealis clauses is a common error among foreigners. Our differentiation of was/were gonna and would in direct speech and in quoted speech is discussed by me in Variation in the data.

     [For the linguist: Some markedness theories--those lacking the theory of reversals in marked categories or contexts--would have a problem (and rightly so) with more subcategories in a marked category than in the corresponding unmarked one--the sort of problem that they have with the fact that Marshallese has got gender in the four plurals but not in the unmarked singular. (W. Mayerthaler has dealt with instances in which the plural is the unmarked variant; but that is not true of this example.)]

     (Notice that true-present is being is not appropriate for exochronous is in the preceding sentence. A true present is heard in "The child is being good right now," whereas "The child is good" is more permanent; Spanish and Portugese make a similar distinction with estar and ser. English, but not other languages, can say without contradiction: "She comes from Dallas, but she 's coming from Atlanta" and "He speaks German but he's speaking English right now." Notice that on Saturday night when the factory is closed, makes cannot represent "present" time in "This factory makes watches"!)

     Now all of this can be treated as a list of data in Mickey-Mouse fashion; or one can look for a SYSTEM that ties it all together, i.e. for a SINGLE principle that accounts for ALL of the foregoing. You would waste your time if you went to the grammar boox. The clue lies in the example where the neutral or default case is called "unmarked"--negation, real interrogation, and comparison (not illustrated) being "marked" as special in some way. The special nature of the category or situation in which forces get switched is also "markered" by some formal differentiation--e.g by switching paired forms (like be and get in the formation of passives or will and be gonna in the formation of posteriors).

      A more sensible approach than just making a list of interesting information is the following:

FIRST, you begin with notions that have been shown in advanced analyses to be correct; thus:

If a personal pronoun is preceded or followed by and or or, or if it is followed by an appositive, it is marked as well as being markered by altering (reversing) its "case" form. This explains the first examples. Who and whom are slightly more complex; we differentiate normal, unforeground "the very one who I spoke to" from formal, foregrounded "the very one to whom I spoke"; we also differentiate the latter from what is said in echo/reclamatory questions like "I spoke to who?"

SECONDLY, if a passive form (with auxiliary verb be or get) or a posterior form (be gonna or will) occurs in a marked time category (see above), it is MARKED, changing its aspectuality, force, etc.; it will be MARKERED by switching forms where this strategy is available (i.e. where two forms exist). This takes care of the examples in the second set above. The force of a pair of forms will also switch in marked environments, as illustrated in other examples above.

Both of the foregoing are generalizable in this principle: In MARKED ("fault," i.e. non-default) categories or situations, the force (unmarked/settled notion, expectative/dubitative, epistemic/deontic, or whatever) and the aspectuality of a verb form, the "case" of a personal pronoun, and so on with other categories is predicted to get reversedin marked situations (categories or environments): Where two forms are involved, the changes of markedness, force, aspectuality, or whatever can be MARKERED by switching the forms; this is very evident in the case forms of the personal pronouns! Stated succinctly, the case-form of a marked pronoun or the aspectuality of a passive verb formation or the expectative/desiderative, settled-notion, or deontic/epistemic force of some other verb modality is markered by reversing its forms.** An additional consequence of this principle is that a doubly (or twice-)marked item will be doubly reversed right back to the unmarked use.

Since the third-person singular and the first-person singular are the least marked person-number categories, it's not surprising that in unmarked situations, was is different from more-marked were--though in irrealis conditionals, were is unmarked everywhere. In the following graded series of was/were, a starred *was is not heard in cultivated usage:

--It was here yesterday and still is now. (unmarked)
--If it was/were here now, . . . (marked as irrealis)
--Were (*was) it here now, . . . (marked additionally by if-DELETION and SUBJ-INVERSION)

     With was for an unmarked person-number form and were for other person-number forms in unmarked contexts, compare the parallel contrast between is/am and are (the latter in marked person-number categories). The unmarked category of an imperative (a marked modality) is the second person (singular). There is no ending in most languages; the negative of the plain imperative often changes its modality, especially in the Romance languages. English further markers an already doubly marked prohibitive by inserting do; such so-called DON'TCHU prohibitives (negative imperatives) reverse the passive forms that prevail in simpler prohibitives, the result of this double reversal being an aspectuality-reversal (in the use of be and get) from what is found in DON'TCHU prohibitives to what is found in least-marked, non-negative imperatives! Hope you followed all o' that and will check the matter out! If necessary, consult the fuller discussion in Variation in the data, pp. 86-87, where various examples--along with the volitional force of DON'TCHU-prohibitives--are discussed. (A true non-negative imperative with get is not used; apparent examples that have misled some analysts are discussed in chap. 6 of Essays on time-based linguistic analysis.) Here are a few examples:

--Don't get caught. (non-processual)
--Don't be caught doin' that. (processual)
--Don'tchu get caught doin' that! (processual)
--Don'tchu be caught now! (non-processual)

Reversals of shall and will in early Modern English follow patterns predicted by the theory. The person-number and modal reversals are analysed in chap. 6 of Essays on time-based linguistic analysis; E. Faingold (see reference in the foregoing book) has analysed reversals of shall and will in the marked contexts of negation and interrogation.

     While the convincing rationale for reversal theory cannot be gone into here (see Essays on time-based linguistic analysis), one may be permitted to point out that calmative drugs given to hyperactive children have the reverse effect on normal children, making them hyperactive! Besides not knowing about reversal theory, the grammarians have been grossly misled by incorrectly assuming that the grammar of English is more Germanic than Romance. The truth is otherwise. The case-forms of the English personal and WH-pronouns, as our language is generally spoken, behave in a manner that greatly parallels their French cousins. As for verbs, it is obvious that the English modals do not resemble the Germanic ones (they lack all of the modalities a Germanic modal has got). And while some aspects of the English verb--e.g. the progressive modalities, the be gonna posterior, the surrealis environment, and infinitive-reversal in negative environments--formally resemble the French verb, especially the old French verb, and our exochronous modality often corresponds to the subjunctive in older French, still the English anterior and posterior modalities (and to a certain extent, the exochronous modality) are sui generis. It should now be evident why one cannot be surprised at teachers of long experience who say and demonstrate that they "don't understand the English verb." The same applies to the case-forms of the personal pronouns--and the WH-pronouns--both of which are clearly close to those of French; in children's English, as in less-educated adult usage, they are nearly identical in the usages of the two languages. And that says a lot for those with sufficient insight to understand the implications of the facts.

     It will be useful now to illustrate how meanings switch when cultivated speakers put the adposition (preposition or postposition) at the beginning of a clause (call it "marked order" if you wish) instead of at the end of a clause--just in case the clause in question begins with a WH-word (who, which, where, when, how, why) or relative-pronoun that. (Participially derived adpositions and others discussed by me in various writings have got to stand in their underlying positions, un-pied-piped.) I haven't got time for the jillions of examples found in other of my writings, but these will illustrate the point; for many others, see my Variation in the data.

When've we got to be home by? vs. By how many inches is Sal taller than Liz?
What're you basing that on? vs. On what basis are you predicting that?
What sort of trouble is he in? vs. In which instances did that cause trouble?

     See Variation in the data for space available vs. available space and examples with in/into (like "When the ball flew in the window, Joe flew into a rage") as well as the difference between adverbial clear and clearly, sure (what grammarians call a sentential adverb) and surely, and many other pairs. (This last difference makes the use of surely for sure in "He's surely [= "sure-footedly"?] leaving" as strange as clearly would sound if it replaced clear in "It is clear at the end of the book." An exception is sentence-intial surely, as in "Surely, they cannot be guilty of that"; cf. "They sure cannot be guilty of that.") Most intensifiers lack -ly; a few, like highly and strongly, are waiting for this ancient pattern to generalize itself completely.

     Which grammar should a foreigner look in to find out whether we say between or among with a singular mass noun like stuff? What grammar explains the rationale of dummy-of which English, like French, has in "City of ..., University of ...," etc., as well as after quantifiers and in many other positions? Does your grammar indefensibly proscribe "all OF the books" and "both of the apples"? Does your grammar deal with "How big OF a deal is that? It's not very big OF a deal"? What misinformation does it propound about "off of, out(side) of, inside of," etc.

     Many interesting reversals of stress (in marked morphological formations) and of intonation patterns (note double reversals--in both word order and intonation--in indirect questions and in echo/reclamatory WH-questions!) take place in English to marker marked constructs; see Variation in the data, pp. 68-69, 91-92, 178-179. Failure to understand the nature of some of these has not debilitated the analysis of English to the extent that similar failures have hurt the analysis of English grammar and formative morphology; but failure to notice the nature of the system is of course rather less than commendable. Marked compounds (with their stress pattern reversed from the default pattern) include proper names; phrasal compounds like jack-in-the-box and compounds with old like old maid; twofer compounds like student-prince; and quantifier compounds like twofold, double-header. These have rising stress patterns, in contrast with the falling pattern of unmarked nouns, etc. Verbs of course have rising stress unless they are markered as denominal by having a falling pattern. (See VID, p. 91, for the the amazing reversals of [Bolinger's] examples, a single[-]roll as a [marked] quantifier compound and as a [marked] noun phrase--note that the default patterns of unmarked nominal compounds and of unmarked phrases get switched. And note that men-o'-war and man-o'-wars don't mean the same thing--any more than mice : mouses do!) Note in addition that Tenth Avenue has a rising stress--a double reversal from the default or unmarked falling compound pattern, which is due to the form's being doubly marked as a quantitative compound and a name. But an additional reversal creates a falling pattern in Tenth Street, since Street is defocused here as default information in context. Contrast rising Hickham Field with falling Honolulu Airport--whose explanation is similar to that of Tenth Street. Bolinger has shown that the same thing occurs intonationally in the contrast between phrasal rising Dora's bike and falling Mike's place and (let me note) even, in a context where bikes are being spoken of, Dora's bike. The explanations for the various phenomena are parallel.

     To forestall misunderstandings, it should be stressed that an ordinarily unmarked construct or use of a marked form class, ordinarily expressed with a marked category, can be "reversed" to what would otherwise be a less markered form in order to indicate a marked use. Thus, while the marked categories of cognitive and sensory (perceptual) verbs are unmarked when used in the marked exochronous modality--remaining non-statives all the while--they can be markered with what is otherwise an unmarked progressive modality to express a special (marked) semantic or pragmatic signification. Thus, sees and hears are unmarked, though the exochronous category is marked as such; e.g. She hears the bells now and He hears you have a modal sense of "can, is able to." This is the way one normally uses these verbs. But one can say, I'm hearing you (where the verb means "paying attention to") and They're seeing if it's doable (where the verb means "investigating, finding out"). Consider a verb that is usually stative--seem: In He's seeming to be in charge, the verb means "trying to appear." I reckon and I'm reckoning convey very different senses. One can try one's own examples with suppose, know, think, and believe. Be knowing is a (marked) posterior, as in She'll be knowing the answer by tomorrow; cf. the normal posterior in a (doubly marked) surrealis environment: If/When he knows the answer, . . . (Those who may have supposed that was knowing is a progressive modality, should take note that very shows that knowing is a predicate adjective in They were very knowing.) Now compare They smell gas or It smells like gas with They've managed it; they're smelling like a rose, even though what they did stinks and She feels tired or It feels like velvet with He's feelin' no pain! in reference to one who is drunk. Similar examples with taste contrast It tastes like dough and They're tasting the wine now.

     Think about these for a week and then e-mail me any questions. Meanwhile, remember: There is a great chasm or disconnect between what grammars say and what educated speakers of English do; it's unreal.

     For further details, see the questions raised in "How English grammars have missed the boat: There's more flummoxing in what they say that meets the eye." This is appearing soon in the Festschrift for W. Lehmann, ed. by E. C. Polome. This copyrighted article will also appear on this website in due course, and is destined to be part of a booklet by Orchid Land Publications to appear in the Fall: How it has gone so wrong with English grammars. The questions raised in "How English grammars have missed the boat" are answered here in a longish article bearing the title, "You and your grammar."

SEE FURTHER HERE & HERE & HERE

---------------------------------------------------------------------------------------------------------------------

     *The reader should not be fooled by examples like "He wonders if they're gonna be on time tomorrow"--with a normal (non-surrealis) posterior. In this example, if introduces an indirect question--not a posterior surrealis hypothesis clause like "If she's on time tomorrow, . . ." ("If they're gonna be on time tomorrow" has the force of strong expectation.) Incidentally, does your grammar inform you of Bolinger's principle for distinguishing indirect questions introduced by whether from those introduced by if?

     **Chomsky has been right in viewing "case" in English and the Romance languages as (a) a matter of order and indeed as (b) a matter of forms rather than simply of function.

SEE ALSO HERE & HERE & HERE