1. The Rezian dialects (spoken in the communities of Resia (Rezija) and Ucce in the district of Moggio near Udine in north-east Italy), which I investigated first in 1873 and several times thereafter, are, in my opinion, a separate dialect group, as are the dialects of certain other language areas (especially Slovenian and Serbo-Croatian). The following <Rezian> phenomena, among others, attracted my attention:
At various times I perceived as different the seemingly identical (linguistically speaking) final consonant -t of the words and word groups: pôt “way,” svît “world,” lît “year,” on the one hand, and pêt “five,” dœvat “nine,” dœsat “ten,” šejst “six,” . . . , pêt lît “five years,” dœvat lît “nine years,” dœsat lît “ten years,” šejst lît “six years” . . . , on the other hand. Sometimes I heard the -t in all these words, but more often only in pôt, lît, . . . , but not in pêt, dœvat, . . . i.e., pê, dœva, dœsa, šejs . . . , pê lît, and, less often, in none of these words; i.e., not only pê ... , but also pô, lî, . . . pe li. . . .
What we are dealing with here is obviously not only the different perception of words ending in —t but of all words with a final consonant. (Cf. my “Versuch einer Phonetik der resianischen Dialekte,” Opyt fonetiki rez’janskix govorov [Warsaw-Petersburg-Leipzig, 1875], §§ 89-103).
2. In the Serbo-Croatian dialect of southern Italy (the communities of Acquaviva, Colle-Croci, San Felice Savo, and Montemitro in the district of Molise of Campobasso province), I discovered a similar phenomenon affecting the final vowels. Thus, for example, the word udovica “widow” is sometimes pronounced udovica, but more frequently udovic. This is evidently connected with the fact that in the Slavic dialect of Molise the importance of the desinential endings as syntactic morphemes has sharply declined under the influence of the neighboring Albanian and Italian dialects and especially the Italian literary language, which is used in schools, business, government, the church, military service, cultural affairs, and literature. We see here one of the final stages in the weakening of the morphologization of endings, a manifestation of the tendency to eliminate inflection.
3. Upon a cursory examination one might regard this as a contradiction of the laws of logic: what seems to be psychologically identical sometimes appears externally as a positive unit and sometimes as a zero. This logical and gnoseological puzzle can be solved if one approaches the question from a psychological viewpoint and bears in mind the fact that in one’s own language, i.e., in actual individual linguistic thought, we do not encounter strictly phonetic or phonetic-acoustic phenomena, but psychologically live psychophonetic phenomena, images, and psychological-social processes. In linguistic thought there are no sounds; there are only representations (concepts) of sounds. But in linguistic intercourse (i.e., in collective linguistic behavior), there are not only linguistic concepts in the individual soul or brain, but the speaker also informs the hearer, by physical means, that he has mobilized at a given moment some of his linguistic concepts, while the hearer receives the impressions and sensations thereby formed. Not everything that is either consciously or semi-consciously dormant in linguistic thought manifests itself every time.
4. Let us consider, for example, the case cited at the beginning of this article: the pronunciation or nonpronunciation of consonants at the end of a word in the Rezian dialect (§1). What factors are here to be considered operating and operative?
First, there is the general psychophonetic tendency to an ever greater weakening of consonants at the end of a word, a tendency which gradually must lead to the total psychological loss of these consonants (more precisely, to their not being reestablished in the minds of a given generation).
Second, there are various degrees of morphologization and semasiologization, various degrees of psychological stress (psychological accent) of individual morphologized and semasiologized phonemes.
Third, there are differences of tempo and style of speech. On the one hand, though more rarely, there is a greater or lesser degree of solemnity and carefulness; on the other hand, in everyday life there is a greater or lesser degree of carelessness. With increasing carelessness in pronunciation, there is a decline in the ability to resist weakening and change. A well-known example is the pronunciation or nonpronunciation of the French e muet.
5. The above-mentioned tendencies toward the weakening and ultimate disappearance of final consonants in the speech of a community is only a special case of the general tendency toward abbreviation, simplification, and relaxation of effort which is present in individual and collective linguistic behavior.
Just as two seemingly identical pieces of iron, glass, or sealing wax behave differently, depending on whether they do or do not possess magnetic or electric energy, so, too, the relative ability of seemingly identical phonemes to withstand quantitative or even qualitative change depends on the degree of their semasiologization and morphologization.
In the Rezian (cf. § 1) indeclinables pêt “five,” dœvat “nine,” the -t is only the final consonant of a single word used in the “general case” (casus generalis, casus unicus), and it plays no special morphological or inflectional role; an understanding of the word in connection with other words hardly requires that its existence be signaled aloud in the communicative process, i.e., that the word be pronounced or perceived as either pêt or pê. Conversely, pôt “way” and lît “year” represent only one case among several others: pôta, pôtu, pôton . . . , lœto, lœton, lîta, lîtan. . . . The final stem consonant -t plays here a highly important morphological role as a morphological link or boundary between two morphemes, and we are constantly reminded of its role. Thus, while the psychological pêt is rendered physically as an audible pê, with pêt being used only in high, solemn speech, the final -t in pôt, lît, resists elimination even in the physically weakened or reduced pronunciation of everyday linguistic intercourse. If, however, carelessness in pronunciation prevails, the psychophonetic tendency toward eventual weakening of the final consonant can cancel out even strong morphologization, so that even pôt, lît, etc., whose final -t is present in the speaker’s mind, are rendered by the shortened sound forms pô, lî. . . .
6. The same applies to all cases of weakening of psychophonetic energy, of lessening of psychological stress, of the decrease in the semasiologization and morphologization of speech elements that takes place with the passage of time in the history of all collective languages without exception. There is a continuous “struggle” between two opposing tendencies: between a progressive drive toward simplification, ease, and elimination of superfluous activities, and a conservative, therapeutic force striving, also physically, to underscore indispensable psychophonetic elements. If one takes his speech activity seriously, accurately, and responsibly and wishes to express and to project into the external world everything that is mobilized in the mind, then the conservative principle prevails. But if one treats his own speech less rigorously, the mobilization of the sound image, or the linguistic representation of the less essential elements of language in the mind will not be strong enough to reach the world of the senses or to become perceptible and to signal its presence. Then only the intention remains, without its psychological realization. When carelessness of speech is increased, i.e., when speech energy is reduced, the tendency toward a psychophonetic weakening of the categories that correspond to the acoustic elements prevails, so that they all remain, regardless of the degree of their semasiologized or morphologized prominence, in a state of psychological mobilization without affecting the world of external perception.
7. There are, of course, individuals who speak very carefully and clearly and whose pronunciation counteracts the weakenings and changes in the language of the community. But such individuals are the exception. Most people speak slovenly. An unaccustomed listener or participant in a conversation “does not get the full message,” is not in a position to understand it, and asks that it be repeated; upon repetition, the intended psychological sound-representation is, as a rule, rendered with precision.
But even individuals who are extremely slovenly in their speech usually write far more accurately, conveying their linguistic concepts and representations more precisely in the sphere of optics than in that of acoustics. Writing thus exerts a conservative effect on speech habits. It often happens that children learning to read conceive their natural speech as faulty and unsuitable, and instead of speaking, begin, in fact, to read. Their speaking is guided by their reading.
8. The weakening and disappearance of phonemes and other psychophonetic elements which are not sufficiently semasiologized and morphologized is counteracted by practice in the solemn and high style of speech which is used in schools, offices, on the stage, on the podium, at official meetings, at public gatherings, etc.
9. The revival or restoration of psychophonetic elements that have seemingly disappeared and can no longer be reproduced in the national language can be effected through communication with representatives of other dialects in which a certain phoneme is still psychologically present and externally manifested.
10. An inhibiting influence on the course of change and a stimulus toward the revival of disappearing elements is, in the case of literate and educated speakers, exerted by the familiarity of these speakers with the earlier stages of the development of the given linguistic forms (material): knowledge of Latin among the French and Italians; knowledge of standard literary French (or of standard Italian or standard Spanish) by those who speak the modern and recent French patois.
11. The use of slovenly speech in everyday life is, nevertheless, far more common than the use of clear and solemn speech; and because of its high frequency, it ultimately prevails. The accretion of the phenomena of slovenly speech in the course of many generations, i.e., the nonpronunciation of weakly equipped phonemes, brings about their extinction. A phoneme approaching extinction conveys to the representatives of one or another generation such weak acoustic impressions of its psychological stress that these no longer suffice to mobilize their psychological attention, i.e., to awaken the old sound image in the mind of the next generation.
The phoneme looses the energy necessary for its existence, and there occurs a historical transition of a certain positive quantity into a historical-phonetic zero.
What was acoustically and physically facultative for an earlier generation ceases to exist even psychologically for a later generation.
12. This is, for example, how we should think of the transition of the final consonants inherited from the Latin linguistic state to a zero in Italian, French, and in all other Romance dialects, i.e., a reduction of consonants characteristic of the later phase of the Italo-Romance linguistic material. . . .