Consider the possibility that English grammar has been misanalysed for centuries because of grammarians’ accepting fundamentally flawed assumptions about grammar and, even more so, about the history of English—and that this has resulted in a huge disconnect between English grammars and the genius of the English that really exists. The development of the information age and of English as a world language means that such lapses have greater import than formerly. But what is available on the shelves has fallen into sufficient discredit that the uses of these paired forms are not directly related to functional (subject and object) "cases" in the manner of Latin and Anglo-Saxon, but only in a very round-about way that reminds one of the Romance languages. Shouldn’t your grammar offer you some hint of an explanation for usages of pronoun forms in the English of educated people—individuals who would never say to I, to she, or for we—but who nevertheless use constructs like the following: to she and I; Let’s you and I do it; to she who knows; on we in Europe; and "for we Europeans"—the last from the mouth of a British prime minister and former university professor? If the educated generally say a rather Frenchified "Martha and myself arrived late" and "That’s me" (Marilyn Monroe was made to sound silly by saying "That’s I!"), why do grammarians ignore the significance of such usages for the system of English? Aren’t they interested in why quasi-French usages like "Me

——————————————————————————————
¹This writing first appeared with a slightly different title in Language change and typological variation: papers in honor of Winfred P. Lehmann on the occasion of his 82nd birthday, eds. C. F. Justus and E. C. Polomé, as part of the Journal of Indo-Euopean studies monograph series, 1998. It is reproduced here by permission of the copyright owner; the format has been slightly altered to fit that of this volume. This is one of an on-going series of non-technical treatments by the author of English grammar underpinned by the theoretical analyses of the author’s Essays on time-based linguistic analysis (Oxford University Presss, 1996; see especially Chh. 6 and 10). The point of departure is a determination to see the categories of English grammar for what they are—not through the incompatible lens of Anglo-Saxon or Latin typology.

and my father cleaned the house up today"—typical of that legendary 13¾-year-old British adolescent, Adrian Mole—are so ineradicable? Shouldn’t a grammar give some hint as to why many educated speakers insist that for her and me is ungrammatical? Grammarians can hardly avoid conceding that many usages violating their prescribed "case" rules (like "Who’ll do it if not me?" and "Someone—probably me—will do it") are in fact quite unexceptionable. To contend that these deeply ingrained and ineradicable Romance-language usages are contrary to the genius of English amounts to no more than the unconvincing arguments of a seriously defective analysis. The system (a very strict one) of real English grammar is missing from the books. An adequate approach might offer some insight into why grammar gurus unwittingly write things contrary to their own prescriptions like "the one whom I had been assured was going to fix it" and "It was for whomever over there wanted to win quick acclaim."
     If your grammar calls makes a "present tense’ in "This factory makes machines"—say, on Saturday night when the place is closed—is that a designation that makes sense? Would it not imply contradictions—which are in fact absent—in "She speaks Japanese but she’s speaking English" and in "He comes from Honolulu but he’s coming from Hilo"? Would a grammar that calls makes, speaks, and comes "presents" be expected to do any better with the two posterior uses of speaks in "The president speaks at ten tonight" and (with a different force) "When she speaks at the meeting tomorrow, . . ."? Treating speaks as a present in "Hamlet speaks to Ophelia in this scene," "St. Paul speaks of that in Galatians," and "She speaks about that every time I go there" conflicts with the obvious. (In "He smokes now," now means ‘nowadays.’) Doesn’t the reader know a language studied by many Americans in which speaks would be a subjunctive in the context found in "Make sure that your report speaks directly to the issue"?
     Does your grammar of choice tell you why the import of was being aimed at (in "It was being aimed at") gets changed to was getting aimed at in "It was getting aimed at for a take-over next month"; and why the import of got shot (in "He got shot yesterday") switches to was being shot in "He was being shot at dawn the next day"? Calling the posterior forms will and be gonna (so written to distinguish it from the literal sense of be going to) "future tenses" leaves one wondering how it is possible to find a future in was gonna leave yesterday, seeing that yesterday is past; and why now sounds all right in "They’ll now be arriving late." These are not minor issues. What "tense" is found in "Those problems had been going to have been getting investigated by now"? In the improbable event that your grammar tells you why—even that—the import of was gonna rain gets changed to would rain in an indirect quotation, does it explain that, let alone why, we negate "It’s gonna rain" with "It won’t rain" and "It’ll rain" with "It’s not gonna rain"?
     But we don’t use either of the foregoing posterior forms (English has got nine in all) in "If the president speaks tomorrow, we’ll know the answer," except in infrequent instances when we have good reason to assume that the posterior event is pretty much of a sure thing. The foreigner’s "When she will speak" sounds bizarre for "When she speaks." (The environment here, technically called the surrealis environment, is different from the plain environment in which speaks in "She speaks tomorrow" indicates a routine or scheduled event or state.) We say "After [or Before] he speaks tomorrow, the committee will meet," "When [or as soon as] she speaks tomorrow, we’ll know the answer," "Till he speaks about it, we won’t know," and "Whoever speaks tonight will be applauded." Does your grammar tell you what is systematically going on here—what behaviors are common to surrealis environments? Does it even tell you why "To speak like that is being considered reprehensible" is un-English—or why the replacement of is being with the class of verb modalities partially represented by is, has been, will be, may be, or might be makes this example quite acceptable?
     Although English does not lack a strict system of grammar, what that system is is far from being transparent in grammars that call has stood a "present perfect" in "That building has stood for ten years [and is likely to last a lot longer]." This action or state is continuing in the present, though it doesn’t have to do so for it to be a present-anterior. The English present-anterior refers to actions or states occurring (without necessarily being concluded) in a block of time that includes the present, whereas a perfect is an event or state "completed" or ended in a block of time prior to the reference time. This explains why people are now saying, "I just did it" (which is more logical than replacing the present-anterior with a past in "I didn’t do it yet"). An adequate grammar should give grammatical grounds for why "As long as I can remember, she’s always done that at midnight" is acceptable, even though "She has done that at midnight" is un-English. Grammars are defective in failing to give any hint of the factors—at least two or three striking behaviors—that are common to the following (irrealis) categories: timeless forms like speaks, anteriors (formed on have), posteriors, imperatives, infinitives, and modal verbs.
     One of the most widely used books on grammar (a fairly recent British example) calls a form like was standing a past-continuous tense—even though this form is strange (the corresponding Romance form is also odd) in continuous past examples like "Troy was standing 600 years." The same grammar achieves the ridiculous in pointing out that this verb form "can also be a way of showing your interest in the other person" and "to show a change of mind"! But can’t all of the other time forms of the verb convey the same thing (cf. "I’d been planning to stay home this evening, but I won’t")? Even if such an approach could go so far as to list all of the usages of every verb form, that would still fall short of being an analysis or a grammar!
     You can test your grammar by checking whether it offers a subjective, ad hoc listing of usages in place of grammatical explanations, say, for the difference between Don’t do it and Don’tchu do it! and between doesn’t need to and needn’t—and whether it tells you which classes of environments needn’t can occur in—and why. Does it explain the grammatical reason why you’ll have to is politer than you must and why you can is politer than you may? Also to be checked is whether your grammar helps foreign teachers of English to see what is so wrong when they say until (for by) in "They will begin it till ten o’clock" —or even "They will kill the hen till ten o’clock"—although "She is not gonna kill the hen till ten o’clock," "He was gonna kill time until ten o’clock," "The opposing armies will kill each other right up till the truce goes into effect," and (an acceptable sense of) "Lock the door till ten o’clock" are all good English!
     Consider the difference between should—which is usually dubitative when not obligative, as in "If we should be late, . . ."—but is factual in "They were surprised and disgusted that he should say such things"—which would not be said unless the person in question really was saying such things. After all, even a child’s Uncle Remus stories are full of examples of factual should; e.g. ‘An’ who should ’e meet but Brer Fox!" A reliable grammar should tell you about the four uses of should—what they are, where they occur, and why they differ. You can test whether your grammatical authority calls the deletion of should in "It was for the good of the country that she not remain" a "subjunctive"—even though the processual mode, should-DELETION, explains three anomalies that no subjunctive, past or present, can account for. (See more on this below.)
     Consider the difference between in and into in "I put my money in my safe, but she put hers into stocks and bonds’ and in ‘When the ball flew in the window, he flew into a rage.’ In such typical examples (though not in every one of the four classes of into-usage), into signals an ontological or situational change—something very different from the change of place found in Germanic usage but not lacking in affinities with French dans and en. (Why do Germans speaking English say "involve into" as well as "evolve into"?) Consider now the difference between prepositional and postpositional by, in, etc. We always prepose by and in in "By what right are they claiming that priority?" (contrast un-English What right . . . by?) or "By how many inches is she taller than me?" (cf. unacceptable How many inches . . . by?), and "In what instances [or ways] are they different?" (contrast un-English What instances [or ways] . . . in?). And so with other prepositions. On the other hand, these examples contrast with the unforegrounded by and in that we naturally prefer in "What time have we got to be home by?" and "What kind of trouble are they in?" And so with other postpositions. Since there is a real difference in these usages—one that is known—grammars ought not to neglect the facts or the grounds for the difference. (Exceptions like during, concerning, except, and according to have a ready explanation that your grammar should tell you about.) Though foregrounding a preposition where we don’t have to do so feels unnatural (and in fact violates an implicational universal long known from work by H. J. Greenberg), it can be done to marker insinuations like a resigned For what? (instead of the usual What for?).  
     Where foregrounding is not used to marker irony, frustration, or other insinuations, it can be used to create a put-down (as in "With whom do you wish to speak?" instead of the affable "Who d’you wanna talk to?") or to avoid syntactic complications, including a succession of two prepositions belonging to different clauses. Just as we use of in deviation from what would be a normal ’s in order to avoid awkward constructs, so we prepose a normal postposition to avoid, say, the awkwardness of "the program that you were determined to involve all of the projects that they came up with *in." Needlessly preposing adpositions is otherwise unnatural enough to trap speakers in a redundancy as old as Malory—like this from a BBC announcer: "the one of whom he took advantage of."
     Even a half-way adequate grammar ought to tell us why putting the preposition at the beginning of its clause is un-English twice over in "In which subjects they excel is about what her report is"; cf. the three misuses of preposing in "We wondered about on what her reports were" and "We wondered about from where they had come to get here." (There is no problem with "We inquired about how many there were.") Though the use of from where? (in direct or indirect questions) is a common error among foreigners, non-interrogative from where is all right in "I flew to Honolulu, from where I proceeded to Hilo by water." It’s too bad that a "Pygmalion"—which is to say, a bookish indeed and alien-sounding—approach requires moving prepositions along with their WH-pronoun objects to the beginning of their clauses even in the unacceptable examples above.
     Consider the difference in the uses of partially similar adverb forms—hard : hardly, most : mostly, very : verily, just : justly, bloody : bloodily, right : rightly, clear : clearly, stark : starkly, pretty : prettily, sure : surely, awful : awfully, etc. Your grammar should explain the difference between "It’s sure concealed" and "It’s surely concealed"—and why it sounds silly to reply to normal requests with "Surely"! Being plain stupid is not the same as being plainly stupid; and being real special is different from being really (and truly) special. Your grammar should also clarify why pretty ugly, bloody neat, and jolly sad are not contradictory.
     And shouldn’t your grammar give you the grammatical reason for why there is no contradiction in "That irresponsible person was the person responsible" or "The space available was less than the available space" . . . and why "He’s no teacher" is said of one who in fact is a teacher, though "She isn’t a teacher" refers to one who really isn’t a teacher? And aren’t these last two examples paralleled by "They’re no better than us" and "They’re not stupid" in that no better can imply that they and we are not very good, whereas not stupid means just what it says?
     Does your grammar help you understand the error in an editorial of a prestigious newspaper—the use of were instead of expected was or had been —in "Mr. Starr should not have taken it unless he were willing to see it through"? And how about the may in "If they would’ve [for had] installed a smoke detector, this disaster may not have happened"—which makes even less sense than the televangelist’s might in "Let us pray that this child might receive healing"—itself pretty bizarre? When (in speaking) we use a past counterfactual if-clause, we often strengthen ’d (which gets assimilated and deleted in conversational English); we change it to had’ve, as in "If they’da [or hadda] been there"; the British playwright, D. M. Storey writes "had have" in Home). This is something for which the grounds are both teleological-theoretical (technically: constructional iconism) and pronunciational; your grammar ought to account for both. (Both French and German "surcomposite" verb forms have different uses from the uses of their formal parallels in English.)
     You can test your grammar’s logic by checking whether it sets up an entire category—a mythical "subjunctive mode," on the basis of a single and (usually optional) re-spelling—viz. of was to were; e.g. in "I wish she were/was here." (The re-spelling is obligatory in "Were she here now, . . .," but is of course not appropriate in "If Joe was at home yesterday, . . ."—an open conditional). You can test your grammar’s logic also by checking whether it calls the postposition ’s in the person I spoke to’s home town a "possessive" (does she "possess" her home town?) and even a case ending (if so, on what word?).
     Don’t readers of grammars sometimes wonder why grammarians force English into an alien mould? Wasn’t the dummy-of in outside of (which simply calques French hors de and is quite consistent with the English system) formerly denounced simply for not being Germanic (Anglo-Saxon)? Doesn’t a quick look at the small difference between comparatives and superlatives in French show why English might prefer the superlative to the comparative in "That’s the best of the two"? Grammarians seem to think that English is a direct descendent of Anglo-Saxon because both were spoken on the same territory, though such reasoning would mean that English in Ireland descends from Gaelic!
     The basic failing stems from relying on lists of building blocks—easily borrowed words—instead of the unborrowable architecture of the language. It stems equally from treating calques, especially functors, on the basis of their derivation instead of on the basis of their functioning in the system. Like all new languages a-borning, Middle English got its words wherever it could. Its system, however, grew out of the language spoken by all of the bishops and all but one of the nobility of England; it simply defies all sociolinguistic credibility to assume that the language of the underclass should have (except with respect to intonation) prevailed over that of the gentry. The language of the Norman rulers bears a relation to Middle English that offers parallels with the way modern English is related to her many daughter languages in Africa, the Caribbean, and the Pacific.
     Even amid the current re-evaluation of what English grammar is, many grammarians lamely resort to speaking of "standard" English, as though there were an academy (something rejected by Samuel Johnson as un-English) to standardize it, as though the editors of two of the best-known dictionaries in England subsequent to Johnson’s had not rejected the concept of a standard English! While it is in quite order to protest against usages that wantonly make the grammar inconsistent, it is also in order to accept usages that accord with the genius of a language. (How can they be wrong in English if they are right in other languages; e.g. French?) It is thus a vain and losing battle to contend against natural processes like making lay a contraponent verb (as most people do) or against making loan a causative verb (as is also widespread).
     Processual modalities are naturally ruled out of the prevailing static-reic approach to grammar. A good example is the TEMPORAL THROW-BACK in "If they were here now, . . ." and in "If they had been here yesterday, . . ." This processual modality makes a statement less assertive; in the surrealis environment, it conveys counterfactual force in real past or present time, dubitative or counterexpectative force in posterior time. To treat the throwing back of time as a thing—viz. a misnamed "subjunctive mode"—evinces an untenable, static philosophy of grammar. (English grammar includes eight processual verb modalities.)
     When grammarians leave aside lists of unconnected items and concentrate on unborrowable systematic structures, they will abandon that mythical "subjunctive" that frightens school children and needlessly adds to the inventory of items that they’ve got to master. Consider the example given earlier: ‘It was for the good of the country that she not remain.’ There are at least four anomalies here, including the lack of a sequencing of time forms (which an alleged subjunctive cannot explain, inasmuch as the long defunct subjunctive obeyed this principle), the absence of do-SUPPORT with not, the lack of an inflection on remain (the only part of the matter that the former subjunctive might account for)—and the optional use of should before remain. Indicating the processual modality of should-DELETION with the notation S—which plays the same rôle (e.g. that of having an infinitive complement) as the deleted should would—explains the construct without further ado. Its function is to preserve the structure. All that one now has got to do—it is necessary in any analysis—is to observe that the construct occurs after expressions of necessity and will and those of propriety and importance.
     A similar technique eliminates the proliferation of unnecessary building blocks—including contradictions like "adverbial nouns" and "substantival adjectives." So-called adverbial nouns are really adverbial prepositional phrases with deleted prepositions, indicated by P in "They did it P my way"; most are measurements—as in "We worked P eight hours yesterday." The P P functions just like the deleted preposition; in languages with nominals inflected for case, it will govern the case of its object nominal. As for alleged "substantival adjectives," consider the good in "The good O is what we seek" and in "The good O do not always have the best luck"—where O represents deleted one and ones or quasipronominal thing and people. Since the O retains the properties of the deleted nominal, the difference between is and are in the examples just given shows that a predicate will agree with the grammatical number of a subject O! Such simple devices preserve grammatical structure and eliminate unnecessary and unsystematic elements. Additional examples and details are found in the longish booklet mentioned in the footnote, including the X that provides the explanation for postposing the adjective in person responsible X and space available X (in both of which X = "for that [purpose]") as well as in the goals already achieved X (where X = "by someone/something"). (This will be evident to anyone cognizant of the non-Germanic rule [the English examples have got affinities with French une table longue de trois mètres in its contrast with une très longue table] for postposing adjectival phrases after the nouns they modify. X preserves the structure that triggers the rule in question, being structurally equivalent to the that-clause and infinitive construct that respectively modify happy in the débutante happy that she had been invited and the débutante happy to have been invited.)
     The grammar is made more rational also with rules like not-RAISING—which explains why English has got "I don’t think that it happened like that" for "I think that it didn’t happen like that." The raising of not in fact explains more, e.g. why "We didn’t think that many attended" means "We thought that few attended"—given that few = "not many." (Only may also be raised; it is treated as a negative by other grammatical rules.) Consider raising also in mustn’t stay. It is no trivial matter that, unlike German muss nicht bleiben (and English needn’t stay)—but like French il ne faut pas rester—the negative enclitic on mustn’t negates the complementary infinitive rather than the main verb itself. In short, not is "raised" from the modal verb’s complement to become attached to the modal itself. Whether the reader agrees or not that the sense ("forbid") of the French and English expressions reveals an unborrowable systematic element common to both languages, all that has been said makes the assumption that English is analysable in terms of Anglo-Saxon categories highly counterintuitive. If one can bring oneself to believe that mustn’t is part of the French system (and many other systematic factors of a compelling nature discussed in the writings cited in the footnote) and could not get borrowed into another language the way words and surficial word order get borrowed, that will constrain and mould one’s view of the ancestry of English. When grammarians find out the true nature of the English grammatical system, a new dawn will have brought an end to simply listing usages—not least to listing supernumerary and contradictory elements—and not to speak of including grammatical categories that belong to English to the same degree that fins belong on an elephant or a trunk belongs on a trout.