- Ernst Pulgram, “Phoneme and Grapheme: A Parallel,” Word VII (1951), 15–20; now also available in the Bobbs-Merrill Reprint Series in Linguistics, Language No. 76 (1964). Josef Vachek, “Two Chapters on Written English,” Brno Studies in English I (1959), 7–34, says: “The parallelism of phonemes and graphemes was consistently, if not always quite adequately, developed by E. Pulgram [1951].” (Fn. 17). The present paper attempts a wider development in terms of a general theory of writing.
- The last are sometimes called logograms, and distinguished from ideograms proper, which are objective illustrations of something. Since ideograms easily become stylized and equally non-objective representations of things, the difference between them and logograms is at times difficult to state, and in any event not necessary for my present purposes.
- Karl Bühler, Sprachtheorie (Jena, 1934); the translation is by Paul L. Garvin, in his obituary for Bühler in Language XL (1964), 633–634.
- I borrow this term (though in my source it is used in a narrower context) from Charles F. Hockett, “Linguistic Elements and their Relation,” Language XXXVII (1961), 29–53, 52: “… morphemes are not composed of, but only (in part) attested by phonemes…”
- This attitude may be disparaged by some as “mentalistic.” It is. It makes little sense, in my view, to exorcise by onomancy the word “mind” and to pretend that what it stands for either does not exist or should not be talked about, while at the same time we cannot help observing, and operating with, the palpable effects of some force—which might as well be called “mind,” little though we know about its locus or its nature; it makes little sense to banish reference to the very agency which caused man to become the only talking animal. Something there is, and it is manifest in that humans possess culture and language; and we cannot but take cognizance of it and assign it a place in our theories about culture and language. “The Bloomfieldian [taxonomic, anti-mentalistic] mode of interpreting features of linguistic description is like contending that the pressure of a gas on the walls of its container is the effect of molecules striking the walls, and at the same time denying that a molecule is a real physical object.” And I agree that “only by introducing mentalistic concepts into our theories do we provide ourselves with the conceptual machinery which makes it possible to account for the full range of linguistic facts.” (Jerrold J. Katz, “Mentalism in Linguistics,” Language XL [1964], 136 and 127, respectively.)
- Cf. Pulgram 1951, 18–19.
- Kenneth L. Pike, “Operational Phonemics in Reference to Linguistic Relativity,” Journal of the Acoustical Society of America XXIV (1952), 619.
- It is worth noting that non-segmental (suprasegmental, prosodie) features which in English phonology pertain to the phonological word, which indeed serve to define and delimit it, function in analogous manner for the French phonological word, which is the breath group. (Cf. Ernst Pulgram, “Prosodie Systems: French,” Lingua XIII [1965], 125–144.) What appears as different in the behavior of English and French prosodies is therefore an illusion that derives from the fact that the fairly consistent coincidence of graphic and phonic word in English is not matched in French.
- Roman Jakobson and Morris Halle, Fundamentals of Language (The Hague, 1956).
- C. E. Bazell, “The Grapheme,” Litera III (1956), 43–46, took exception to my 1951 article (see above, fn. 1) on the grounds that the grapheme corresponds, not to the phoneme, but to the morpheme; the graphic details paralleling the phoneme, he thought, are what I just called the distinctive features, while the written word is the equivalent of the whole spoken sentence. Bazell says: “The graphic categories, as compared with the phonic categories, are shifted each time one unit along the hierarchy. It was a mistake to suppose that they occupied the same position in the hierarchy as the units they stand for.” (45) Perhaps my analysis of graphemes into distinctive features, and the consideration of graphemes other than the letters of the alphabet, some of which are indeed signs and analogous to morphemes, may clarify that I did not mean that graphemes “stand for” phonemes. That a grapheme, or several graphemes, may be translatable into a phoneme is irrelevant in the theory of graphic systems as such. The occasional or even frequent matching of graphemes with identifiable phonic counterparts is but coincidental, even if not astonishing in those graphic systems which attest language, for they actually originate with phonic systems fulfilling exactly the same task of communication. This is, I believe, also the meaning of Josef Vachek, “Zum Problem der geschriebenen Sprache,” Travaux du Cercle linguistique de Prague VIII (1939), 94–104, as quoted below in extenso on p. 219. See also id., “Some Remarks on Writing and Phonetic Transcription,” Acta linguistica V (1945–1949), 86–93, and H. J. Uldall, “Speech and Writing,” Acta linguistica IV (1944), 11–16. See also below, p. 219.
- The digraph iu in Old French niule, Modern French nulle /nyl/ ‘not any, none’, that occurs in one of the earliest written French texts, the poem about S. Eulalia (ca. 881), without doubt represents the phoneme /y/, which in Modern French is spelled u, while in the same document the grapheme u renders the phoneme /u/, now spelled ou. Neither of the two orthographies has inherited from the Latin alphabet a distinct grapheme for /y/, and both use a makeshift device, though not the same, to distinguish /y/ from /u/. It is curious, and bears witness to the phonetic acumen and the spelling skill of the writer of the poem, that he chose iu for /y/, a digraph which incorporates precisely those distinctive features, frontness and roundedness, which in his alphabet must be expressed by iand u separately, and which, incidentally, nicely illustrate a non-native pronunciation of French /y/.
- Bertil Malmberg, Structural Linguistics and Human Communications. An Introduction into the Mechanism of Language and the Methodology of Linguistics (Berlin, 1963), p. 15. Since phonemes are meaningless units, it has even been suggested that they are therefore not linguistic entities at all; cf. Ingrid Dal, “Phonologie und Sprachwissenschaft,” Studia Linguistica IV (1950), passim, and 13: “Die Elemente, die ohne Symbolcharakter sind, die ausschliesslich der lautlichen Ebene der Sprache angehören, sind von diesem Standpunkt aus nicht mehr Bestandteile der Sprache.” Apart from the self- contradictory phrasing, the view that only what is meaningful forms properly part of language, whence all figurae are by definition excluded from language (and linguistics?), is surely too restrictive. The same banishment strikes also the syllable (of which I shall speak below) since it, too, is a figura, hence has no “Symbolcharakter” (4).
- A fairly strong circumstantial case can be made for the heuristic (and perhaps even the historic) primacy of the open syllable, that is, for analysing any language into as many open syllables as possible, and for positing closed syllables only where distributional and prosodie rules (phonotactics of initial and final consonants and consonant clusters, presence or absence of phonological morpheme boundaries within an utterance bounded by pauses, etc.) require them: see my brief remarks in the Proceedings of the Fifth International Congress of Phonetic Sciences, Münster i. W., 1964 (in press), and my article cited in fn. 8, above. A tendency toward opening of syllables has been observed in the history of many languages. A more explicit and extensive paper on the phonology of the syllable will be forthcoming.
- The simplest and most efficient of all syllabaries would be one for a language that has but one vowel phoneme and few, or no, consonant clusters: the vowel, being merely a stretch of non-significant voice between consonants, need not be written at all, and the syllabary is practically an alphabet of consonants and consonant clusters.
- It should be added, to avoid misunderstanding, that those speakers of Greek who came to use alphabetic writing lived on the Greek mainland and in Anatolia, centuries 2—w. later than the Cretan and Peloponnesan writers of Minoan B, and probably unaware of their literate predecessors since there appears to have intervened a long period of illiteracy in the entire Aegean region. Hence one ought not to think that the users of the alphabet had tried out the syllabary and, finding it unsatisfactory, discarded it; they had most likely never known it: cf. Ernst Pulgram, “Linear B, Greek, and the Greeks, Glotta XXXVIII (1960), 171–181. Not even a post-Minoan syllabary, of Cyprus, is derived from Minoan B, but is an independent invention: cf. Johannes Friedrich, “Zur schriftgeschichtlichen Wertung der kretischen Linearschrift B,” Minos IV (1956), 6–10.
- Bertil Malmberg, “Voyelle, consonne, syllabe, mot,” Miscelánea… Martinet, III (La Laguna, 1962), 85, speaks of Chinese writing, traffic lights, and other visual signs as exhibiting (in Hjelmslev's terms) “identité entre contenu et expression” in which “la chaîne n'est pas analysable en figures.”
- Johannes Friedrich, “Noch eine moderne Parallele zu den alten Schrifterfindungen. Eine Schrifterfindung bei den Alaska Eskimos,” Zeitschrift der Deutschen Morgenländischen Gesellschaft XCV (1941), 374–414, reports on the invention and development of a writing system by an Eskimo who, albeit not in complete ignorance of alphabetic writing, over the years progressed, in steady search for improvement, from a pictographic over a syllabic to an alphabetic system for writing his dialect.
- It is not impossible that in some societies pictographic writing was preferred precisely because it was difficult, so that only the initiated and the worthy—the priests, the learned, the scribes—were able to master this mystic and enigmatic art, which bordered on (and at times was made to serve) sorcery. In other cultural climates (the feudal of mediaeval Europe and perhaps of Mycenaean Greece) writing may, on the other hand, have been disdained as a tricky, low, but useful aptitude that was best left to one's inferiors, to scribes and to weak, bookish persons, to singers and poets, and that was a superfluous accomplishment and an unfitting occupation for men of action and power.
- The Ogam script of pre-Christian Scotland, Wales, and Ireland has symbols consisting of one to five vertical or slanted strokes each, oriented in four different ways with reference to a base line, giving twenty distinct graphemes—all figurae, of course. One might not call it an alphabet, because of its appearance, but it is in effect, like the Morse code, a re-coding of the Latin alphabet. The same may be said, mutatis mutandis, of a rather recent recoding of the Latin alphabet, namely, shorthand writing in its various forms. It is mainly based on a simplification of the alphabet: the graphic patterns derive their value from the direction of the strokes and loops and from their cursive combinations (where transitions are used distinctively) more than from their shapes (which are plain and sparing); vowels are not written where their omission does not veil the sense; and there is available a store of proper logograms, or ideograms, that is, abbreviations which are signs rather than figurae and which in fact may stand for more than just one sign of the spoken language, hence attest entire spoken phrases. Thus the newest and most efficient graphic system has adopted to advantage some features of the most ancient.
- Vachek 1945/49, 90.
- Vachek 1945/49, 87.
- Vachek 1945/49, 90.
- Uldall 1944, 16. But see fn. 6, above.
- Vachek 1939, 100.
- Linguists who reduce hitherto unwritten languages to writing sometimes wonder whether for a given language they should in their newly devised writing system spell out all phonemic distinctions (including possibly the suprasegmental ones) and even some non-phonemic but important allophonic traits. The question, it seems to me, is badly put. If the linguist is devising an orthography, he has a great number of choices open to him, as long as the product is a reasonably efficient system, preferably, to be sure, an alphabetic one; for if it is assumed that the texts will be read by persons who know the language already, various orthographies can do the job. (English, for example, c..Id b. sp.ll.d w.th..t v.w.ls.nd st. 11 b. r..d w.1l.n., gh, especially if one uses dots in the place of the vowels; if not, th jb bcms mr dffclt bt nt mpssbl; but if one leaves out the consonants a….i.e. o..y..e.o.e.. i. a…oa..e. i..o..i.i.i.y [and writes only the vowels it approaches impossibility], especially i oe ao oi e u [especially if one also omits the puncts]. Classical Latin did not indicate in its spelling the phonemic quantity of the vowels; a few sporadic attempts at reform in this respect by writing long vowels with taller or geminated graphemes proved abortive. Modern Latin texts that use the macron above the long vowel in fact misspell Latin.) If, on the other hand, the linguist commits himself, for good reasons, to an orthography that is completely phonemic, then questions on how to spell do not arise because the orthography is provided by the phonemic analysis. But the very fact that competent linguists sometimes concern themselves with the problem implies that they are not really certain whether the most desirable and most useful orthography is a strictly phonemic one. Cf. Vachek 1959, 1.
- That M stands for mille ‘thousand’ and C for centum ‘hundred’ is accidental, and although the graphemes were no doubt by most Romans understood acrostically, this is not their origin; indeed all the other symbols (l'unum’, V'quinque’, X'decem, L ‘quinquaginta’, D ‘quingenti’) exhibit no such feature. C is the remnant of Greek theta, ⊙, a grapheme for which the Romans, on borrowing the Greek alphabet (probably by way of the Etruscans), found no alphabetic use since they had no corresponding phoneme /th/. The same is true of X, which is the Greek chi /kh/; of L, which came by way of ⊥(bottom); from Greek psi, ↓ /ps/; of M, which came by way of ⊥ from Greek phi, φ/h (with another development leading to a pseudo-alphabetic CIƆ), and of which D is the right half, as indeed V is the upper half of X. (That V is the schematic drawing of a spread five-fingered hand, and X of two such hands, is a pretty but spurious theory.)
- It may be added that this difference in kind between the two systems, enhanced by the Arabic figura for ‘nothing’, the zero grapheme (which the Romans did not know and which also in Arabic is a later addition), shows itself also in the greater efficiency of the figura system for performing arithmetic calculations graphically, rather than on one's fingers or on the abacus.
- See, for example, the trees constructed for French in Malmberg 1963, Figs. 49 and 52.
- Choose new content alerts to be informed about new research of interest to you
- Easy remote access to your institution's subscriptions on any device, from any location
- Save your searches and schedule alerts to send you new results
- Export your search results into a .csv file to support your research