The ambivalence of Canadian English spelling, which recognizes both British orthodoxy and Noah Webster’s innovations, is one of the many indicators of Canada’s role as a half-way house between Europe and the United States. Unlike the latter, which cut itself off from its European roots and has not infrequently drifted towards isolationism, Canada has maintained strong links with both Britain and France. This situation is reflected in the study and teaching of linguistics in Canadian universities. Stemming from the British, French, and American traditions, there is considerable diversity of theoretical position, which can be a great strength insofar as it leads to a dialectic, to a tolerance of others’ prejudices and a willingness to smile at one’s own. The dialectic—the push and pull of argument between different points of view—can also lead to a better understanding of the various schools of linguistics (including one’s own), and how they differ.

Plusieurs observateurs ont noté que le français québécois faisait en réalisation phonétique un plus grand usage de la durée vocalique que le français standard, par exemple, et surtout qu’il utilise la diphtongaison comme mode de réalisation privilégié de la durée dans certains contextes. La présente étude se propose d’en décrire les manifestations et d’en montrer la portée exacte à partir d’exemples pris essentiellement au français populaire de Montréal: au-delà de ses caractéristiques de détail particulières, c’est un bon représentant des grands traits phonétiques du français québécois, à l’exclusion des variétés d’origine acadienne qu’on peut rencontrer dans la Basse Matapédia et la Baie des Chaleurs, aux Iles de la Madeleine et sur la Basse Côte Nord.

A word about the title—Wally (as he was known to most of us) was vitally interested in the English language, including, of course, Old English, which he taught. So it seemed appropriate to use Old English words as attributes to his memory. Perhaps this whole tribute should be in Old English, or, more appropriately, in one of the rarer dialects of Canadian English. But I shall be more realistic. All of you knew Wally personally and/or by reputation. You can fill in gaps from your own remembrance of him.

Gōd Wīs Secg: ‘He wæs gīd’ in all senses of good—a good scholar, a good colleague, a good husband and father, a good friend, a good Canadian, a good citizen of the world. Secg? ‘Secg’ is an interesting word because in the masculine form it means ‘man, warrior, retainer,’ and other things, but in the feminine it means ‘sword.’ Secg is glossed as a poetic word, so I chose it in preference to, say, wer, or man; for Wally had a kind of poetic quality. Formidable in appearance and in scholarship—a fighter, a linguistic swordsman who could be overwhelming—he was essentially a kind and gentle person. Never one for self-praise, he carried his learning lightly and, like a retainer or servant, he was ever ready to share his knowledge and to assist the rest of us.

With a few important exceptions, such as Berko (1958) and Derwing and Baker (1977), linguistic morphology has always been a descriptive science, content with investigating actually occurring words and extracting patterns from these. Consequently, it has suffered from a difficulty common to all descriptive endeavours, the fear of obscuring individual cases in abstracting to general principles. In turn, as a result of this deficiency, morphological theory has tended to expend its energy not on general patterns, but rather on single words: more thought has been devoted to the one English word sang than to most other topics. In an earlier paper, Aronoff and Schvaneveldt (1978) showed how an experimental technique borrowed from cognitive psychology, the Lexical Decision Task, could be used to investigate morphological patterns without focusing primary attention on single actually occurring words. The essence of the technique is to ask native speakers of a language (English in this case) whether certain stimuli are words of their language or not.

Ces lignes de Louis Havet, qui datent de 1919, posent, on ne peut plus clairement, le problème dont nous nous proposons de traiter. Quoique G. Guillaume se soit beaucoup intéressé, surtout dans ses Leçons de linguistique des années 1938–39 et 1949–50B à la question du genre en français, il n’a jamais abordé directement le sujet particulier auquel l’invitait Louis Havet. Il nous semble malgré tout, ce que nous espérons montrer, que ce problème peut se traiter à partir des oppositions fondamentales posées par G. Guillaume entre, d’une part, langue et discours, et, d’autre part, psychisme et sémiologie.

New words can be formed by adding suffixes to other words. Derived words formed in this way may be phonologically transparent with respect to their base word, or they may be opaque; monstrous is preserved in monstrous#ness but not in monstro+sity. The juncture between suffix and stem is either a word boundary (#) or a formative boundary (+), and while word boundary derivations are always transparent, formative boundary derivations usually result in stress shifting to a syllable other than the syllable which is stressed in the base word, vowel quality changing, etc.

Walter S. Avis was so thoroughly a Canadian that it is perhaps necessary to take special note of his contributions to the international community of scholars devoted to the study of English and the English-speaking peoples. Certainly his broad perspective is everywhere revealed in his scholarship. In his study of Canadian eh?, for instance, he speaks of the corpus he collected when his reading was “arbitrarily limited to books in my own library, to newspapers and periodicals that passed normally through my hands, to radio and TV programs, and to such oral examples as I had the opportunity to observe and set down” (1978:174). Yet such “arbitrary limits” encompassed writers and speakers who represent Britain, Canada, the United States, South Africa, and Australia and covered uses from the eighteenth century to the present. With the best of scholars, Avis was meticulous, thorough, wide-ranging, and devoted to the real evidence of real people speaking and writing.

Le corpus de la présente étude a été établi à partir de 145 enquêtes sur le terrain, réalisées entre septembre 1971 et juin 1972 dans 18 paroisses de caractère francophone, établies avant 1875 et situées dans une zone circulaire d’environ 65 milles qui a pour centre géographique la ville de Sherbrooke. Le choix dés informateurs s’est fait à partir d’une liste dressée avec l’aide du curé ou du secrétaire de la municipalité, ou d’un(e) président(e) de club ou de cercle de l’endroit: l’informateur devait être francophone, né de parents qui habitaient la paroisse et dont au moins l’un des deux y était lui-même né, il devait être âgé de 20 à 65 ans et disposé â recevoir les deux enquêteurs après un premier contact par lettre puis par téléphone et qui exposait très brièvement le but de la visite.

Une des difficultés majeures qu’est appelée à rencontrer sur sa route une linguistique visant à expliquer le langage en tant que phénomène est d’apprendre à opérer partout, dans ses analyses, la répartition exacte du possible et du réel. C’est, du reste, à leur impuissance à vaincre cette difficulté, qu’il faut imputer les insuccès répétés, à date ancienne, de la grammaire générale, et, à date plus récente, de la linguistique curieusement appelée ‘structurale.’ Comme si toute connaissance scientifique d’un phénomène quelconque—ce phénomène fût-il le langage humain—ne devait pas être, par nécessité, structurale! A-t-on jamais éprouvé le besoin, du côté des sciences de la nature ou de leur indispensable instrument d’investigation que sont les mathématiques, de parler d’une ‘physique structurale,’ d’une ‘chimie structurale,’ d’une ‘biologie structurale,’ ou encore d’une ‘mathématique structurale’? Tant il est évident aux yeux des praticiens de l’un ou l’autre de ces savoirs établis que l’adjonction à leur désignation de l’épithète en cause constituerait plus qu’une simple redondance: une véritable tautologie.

It is always gratifying when one’s research elicits a response, particularly the detailed discussions by Singh (1980) and Morin (1980) of Walker (1978). It is also clear that both of these authors have contributed to our understanding of the process of epen-thesis in Old French, although I am not convinced that we may yet be confident of having dealt with the process in a fully satisfactory way. The following comments on Singh’s and Morin’s proposals may illustrate certain of the difficulties which remain.

An aspect of the unravelling debate between proponents of ‘abstract’ phonology and ‘natural’ phonology focuses on the role of diachrony in justifying one’s descriptions and in underpinning one’s theories (insofar as they are meant to explain and predict). As a phonetologist who has no theoretical predilections for either transformational generative phonology (TGP) or Hooper’s natural generative phonology (NGP), my viewpoint leads me to view some of the issues raised in Piggott 1980, and in Hooper 1980 which replies to it, as either side-stepping what seem to me to be more fundamental issues or as addressing the issues discussed obliquely instead of head on. Although the fundamental position of developmental linguistics (or lectology) has been clarified at length in other writings, there may be some profit in making those points concrete in direct connection with an actual debate over an actual analysis. Piggott argues against the adequacy of Hooper’s NGP with data taken from Algonquin-Ojibwa. He contrasts Ojibwa nouns that exhibit the deletion of word-final glide plus lax vowel (e.g. anini ‘man’ vs aniniwag ‘men’) and Ojibwa verbs that exhibit the loss of word-final lax vowel but not a preceding glide (e.g. ni-wīndigōw ‘I am a Windigo’ vs ni-wīndigōwi-min ‘we are Windigos’).

Many of the hours I passed learning from Wally Avis were spent discussing practical problems of lexicography. Because of this, and because my tribute to him is being published elsewhere (English World-Wide 1:1, 1980), it seems appropriate to honour him here by dealing with one such problem, especially one that no one has satisfactorily solved. I refer to the treatment in monolingual English dictionaries of idioms—those elusive, hard to define, and harder still to classify phrases that appear usually in boldface as numbered or listed subentries.

Perhaps the most common type of idiom so defined is the fuzzy phrasal verb, and Wally was constantly intrigued by the anomalies of such apparently simple examples as start up. We can say, “Let’s start up the car” and “Let’s start the car up”; we can say, “Let’s start up the hill” but not “Let’s start the hill up.”

Katz (1980:1) views the meaning of a sentence and its logical form as identical and believes that this logical form should receive an intensional rather than an extensional interpretation. He sees Chomsky’s proposal to separate semantic interpretation into two parts, a logical form (LF), which is strictly determined by rules of sentence grammar, and whatever aspects of meaning remain, which are determined by discourse rules and possibly by other cognitive faculties, as pernicious. He further objects to what he interprets as Chomsky’s supporting extensionalism. He argues that Chomsky’s grounds for separating LF from other aspects of meaning are not compelling, that LF is itself incoherent, and that the results of making such a separation are harmful to the study of semantics in linguistics and philosophy alike.

In a modest way, I have something in common with Webster, Mencken, and Avis, the three spokesmen for linguistic autonomy whom I am discussing. True, St. Noah of Amherst ascended to that great lexicographical office in the sky fifty-eight years before I was born; and as varied as my dreaming has been, I have never dreamed of meeting him. Nevertheless, we share a common birthday, October 16; and in the line of duty, I have often been entangled with his successors and their critics, especially with the critics of Merriam’s Webster’s Third New International.

With Mencken, my relationship was more direct. I had heard vaguely about The American Language in my South Carolina boyhood, when—in the eyes of the southern Baptists who dominated my community—Mencken metaphorically had horns, hoofs, and a tail. But I never read it till 1937, two years after I acquired a Ph.D. as a Miltonist. I had gone to the Ann Arbor Linguistic Institute, under orders from my college president (a retired four-star general) to improve my teaching and my use of the language.

One of the many aspects of the study of Canadian English which the late Walter S. Avis pioneered was the study and analysis of the literary representation of a Canadian dialect, first in his M.A. thesis, “The Speech of Sam Slick” (1954), subsequently in the “Note on the Speech of Sam Slick” in the introductory matter of R. E. Watters’ Sam Slick Anthology (1969), and the application of that study in the edition he prepared of that text for linguistically naive modern Canadian readers. Like all those who work in Canadian English, those of us working on the Linguistic Survey of the Ottawa Valley owe a great debt to Walter Avis, not only for his outstanding work in the lexicography of Canadian English and his studies of the Loyalist based dialects of Southern Ontario, but also for many more immediate favours and services: his whole-hearted support of our project, his answers to queries about particular lexical items, his visits to the project and the commentaries he made on our tapes and our transcriptions of them and, outstandingly, his willingness to let us borrow tapes which he himself had made of Valley speakers, and to let us use them for our own purposes. The study which follows tries to reflect something of our debt by discussing the literary representation of another Canadian dialect, one which is spoken within the area under investigation by the Linguistic Survey of the Ottawa Valley, and one heavily influenced by the Loyalist dialects of Southern Ontario which feature so largely in Avis’s work. The dialect in question is the dialect of Glengarry County, Ontario, and more particularly the two northern townships of Glengarry County, and the representation of the dialect of that area in the Glengarry novels of Ralph Connor.

One of the conundrums of the historical phonology of Spanish is the seemingly intrusive -s- in the word hasta ‘until.’ Its putative etymon, the Arabic ḥattā ‘up to, until, as far as, even, in order to’ does not have this -s- and, as far as is known, has never had it. The most recent serious discussion of and accounting for this -s- that I am aware of is that presented by Malkiel (1979) in which, dismissing previous lexical and phonological attempts as inadequate, he suggests that hasta (earlier fasta) resulted from a contamination of its predecessor fata, coming unambiguously from ḥattā, with a Romance form fascas ‘nearly, almost,’ derived in part from Latin usque ‘until.’ Both fata and fascas then went out of use, and fasta (> hasta) alone remains. In spite of the wealth of documentation adduced by Malkiel, his evidence adds up to this: before roughly 1300, there was a -word fata ‘until’ and a word fascas ‘almost’; after 1300 both fell out of use, and there was a new word fasta ‘until.’ The rest is speculation.