The object of this article is to illustrate a particular model of grammatical description by using it to describe the Terena phrase. This model can, for convenience, be given the label “structure-function.” Like any model it employs certain interrelated categories and terms which are mutually defining. Each category can only be understood in reference to the model as a whole.

This model sets up grammatical units which are hierarchically arranged. The hierarchy consists of a series of levels of description. At each level there is a grammatical unit and this grammatical unit gives its name to the level concerned. At a given level there is only one grammatical unit. Each such grammatical unit consists of one or more of the units next below it in the hierarchy.

Desservie par une terminologie ambiguë la morphophonologie n’a jusqu’ici intéressé qu’un petit nombre de linguistes qui, au reste, ne sont pas encore parvenus à définir ses limites exactes. Si T. de Felice, E. A. Nida et A. Martinet, entre autres, ont noté l’existence des variations phonologiques au sein du système morphologique, le premier n’a pas poussé ses recherches assez loin pour aboutir à la découverte d’un système et les deux autres semblent avoir jugé un tel travail inutile.

This paper is concerned with so-called demonstratives. Languages use various devices to point out or indicate things; the demonstratives are but one of these devices. Much that has been written on this problem of deixis is confused and misleading because the criteria traditionally used in describing the different devices for pointing out or indicating have been for the most part notional or mentalistic; it behoves the theoretical linguist to seek for formal linguistic criteria to describe this feature, and these notes seek to outline some of the problems which are concerned with a satisfactory linguistic explanation of the grammatical category known as the “demonstratives.” In view of the title of this article, it might also be pointed out en passant (though the author has no wish to involve himself here in a terminological wrangle) that “deixis” may well be a more suitable term for this feature in view of the fact that languages possess other demonstrative devices or indicators in addition to those referred to traditionally as “demonstratives,” not to mention the quasi-linguistic feature of gesture.

While the only Thracian inscription known so far remains the one on the golden ring brought to light in 1912, in Ezerovo, near Philippopolis, the possibility cannot be ruled out of finding other inscriptions in this language in the not too distant future. At least two important discoveries have been made recently on the territory of ancient Thrace: the richly decorated royal tomb, near Kazanlak, dating from the third century B.C., which was found in 1944, and the objects excavated at the site of Seuthopolis, a city which seems to have gone down in flames around 229 B.C., after a life of some 130 years. The site was allowed to disappear at the bottom of a man-made lake, following the completion of a hydro-electric project in 1954. Prior to that, valuable archaeological material had been recovered, including more than 1,100 ancient coins, and an interesting inscription in Greek, concerning relations between Thrace and Macedon.

In 1961 G. Rondeau published in this journal an article entitled “Les unités linguistiques de base chez A. Martinet,” in which his intention was to discuss, “aussi objectivement que possible,” the basic linguistic units as they appear in A. Martinet’s handbook for beginners in linguistics.

Objectivity? A. Martinet states in his preface that certain novel elements are to be found in chapter IV, the chapter dealing with morphology and syntax. And since G. Rondeau’s article discusses Martinet’s principles of morphology rather than the “unités linguistiques” in general, one would have expected that Rondeau would have aimed his arguments mainly against these forty-nine pages of chapter IV. But it appears that the reviewer has been unwilling to extend his interest into this chapter.

Le récent colloque de linguistique française, tenu au Centre international d’études pédagogiques de Sèvres du 20 au 22 décembre dernier, soulignait fort à propos l’unité profonde de la langue, dans laquelle l’analyste découpe des tranches et des niveaux avec l’insouciance d’un chirurgien en salle de dissection. Or, en fait, ces unités et ces niveaux avec lesquels nous opérons sont si bien entremêlés que, pour mieux les voir, nous risquons de détruire la matière même que nous prétendons connaître.

It is strange to observe how many linguists do not know how to exploit an etymology properly. One of the sad consequences of this ignorance is the destructive and unjustifiable rejection of traditional vocabulary within the realm of grammar which linguistics is supposed to defend. The symptom indicates the more serious condition of an ignorance of profound grammatical truths that the ancient Romans and Greeks had discovered, but which, for the most part, have escaped our attention because often we do not know how to use words properly. As an illustration of this point, I ask you to reconsider with me the word “case.” Its grammatical meaning is almost always a source of confusion even though both English and French have cases. Since the word comes to us from Latin, I shall first examine what “case” is and how it works out in the Latin language.

On establishing a Standard Literary Language (SLL). Having once designed a SO to serve all the Canadian Eskimo dialects, the question of selecting a standard literary language remains. It is important to repeat that the existence of the former does not presuppose the use of the latter. The best proof of this is the existence of the syllabary which most Canadian dialects have used as a SO although not one of them is as yet recognized as representing a SLL.

Most languages of long literary tradition developed, at least in the early stages, without any organized central control. The fact that the Canadian government and other groups such as religious bodies can benefit themselves and the Eskimos by publishing their texts in a SLL written phonemically makes it necessary to decide on the criteria that will govern the official choice of the dialect that will form the basis of the SLL. To wait for the natural process of history for the creation of a SLL among a population of some 10,000 Eskimos would involve too long a period of time and might even endanger the survival of their language under the increasing influence of the white man’s language.