Fourth, what shall I say about writing?
(1) Or about its lack (since it is more than likely that when men and language divided and scattered there was as yet no knowledge of alphabetic writing nor of any signs for sounds, if it is the case that the language then spoken was as infantile as we have deemed probable on the basis of its consequences). A lack that deprived words, expressions, meanings of all [1268] fixity, law, form, certainty, or precision, and that left speech fluctuating in the people’s mouths, and subject to their whim, with not the slightest curb or guide or norm. The sheer quantity of variations deriving from such a circumstance may be seen by anyone who cares to observe the dialects into which one and the same spoken language always or almost always divides, even if it is already fully developed and put into written form, and, in short, the countless differences that the language spoken by the people undergoes, depending on time and place, even when it is itself also written, etc. So that, if from what we see we were to make our way back to what must have been in those times, when the ignorance of man was paramount, and paramount, too, were the uncertainty and unsteadiness of all life, etc. etc., we can easily see what either the primitive language or the subprimitive languages must become, and just how many forms they would be bound to assume, since they lacked the support, and the asylum, afforded, not so much by literature, as by alphabetic writing pure and simple.1
(2) What should I say about the invention of writing? Just think, given the early imperfection of this prodigious and profoundly difficult art, the difference between the alphabets, or the impossibility of adapting one people’s written alphabet to fit the spoken alphabet of another, [1269] given the want of expertise of readers and writers and the first scribes, etc. etc., just think what incalculable and unclassifiable alterations the first languages had to undergo, both in their written and their spoken form, when writing began to have an influence on speech.
Note a most remarkable fact. All the ancient languages that have reached us could only have done so by way of writing, for even if they are not wholly dead, the course of the centuries brings such huge variations to languages that who can feel confident in reasoning from the way in which a very old language is spoken now about its ancient properties or even those existing after its formal constitution? Now it is certain that written languages differ to the utmost degree from spoken ones, given the difficulty that was bound to be experienced at the beginning in representing each sound, etc., exactly, a difficulty which invariably produced excessive differences between ancient words as they were written, and as they were pronounced. Such differences gradually became fixed, and despite the care taken on the one hand to arrive at a more precise correspondence between the written signs and the sounds by inventing new signs, etc. etc., despite the influence that scripts would gradually acquire over modifications in speaking, etc., it is certain that such differences, greater in one place, smaller in another, were bound to become perpetual and to be preserved forever.
[1270] Consider, therefore, the dangers we run in reasoning about the properties of an ancient word, and its first form, from the only way in which it can be known to us, from the way, that is, in which it is written. Like someone who reasoned about the English or French, etc., languages from the way in which they are written. There is no rule by which we can know precisely what the value and pronunciation of a specific character in an ancient language was, especially a very ancient one, and especially from the most ancient times, etc. etc. So it is highly probable that very many words from ancient languages—which, when we see them written down, seem to us to be very different and disparate—would seem entirely similar if we only knew the true and original pronunciation that was intended to be represented in the very earliest times by the specific signs that we see. See p. 1283.
Add an observation that reinforces the argument. The invention of the alphabet is so marvelous and difficult that it is highly probable that the first alphabet to be invented passed from the nation and language which invented it to all or almost all the others; and hence that either all or almost all the alphabets derive from a single original alphabet.1 What is sure and certain is that the Phoenician alphabet, the Samaritan, Hebrew, Greek, Arcadian, Pelasgian,2 Etruscan, Latin, Coptic, not to [1271] speak of a good number of others (such as Middle Gothic, Gothic and German, Anglo-Saxon, Russian) obviously demonstrate the unity of their common origin. Yet what languages are more disparate than, e.g., Hebrew and Latin? (Yet they had, as we see, the same alphabet in the beginning.) So much so that Sir W. Jones, who has it that the languages and popular religions of the original stock of the Persians and the Indians, the Romans, the Greeks, the Goths, the ancient Egyptians or Ethiopians derive from one and the same origin, is adamant “that the Jews, the Arabs, the Assyrians, that is to say, the 2nd Persian stock, the peoples who used Syriac, and a numerous tribe of Abyssinians, all spoke another original dialect, and one entirely different from the idiom just mentioned,”1 that is to say, of those other peoples. So that, except for that first nation, where the alphabet was invented, however that came about, all the others, or all which directly or indirectly received it from that nation, wrote using a foreign alphabet. And since the variety of sounds, etc. etc., were infinite among so many nations, see what massive alterations each language had to undergo while they were being adapted to a single alphabet, which, for it, was foreign to a greater or lesser degree, and very often extremely so. See pp. 2012, 2619.
To all of these factors add the far greater alterations the subprimitive languages underwent as they subdivided and subdivided again according to the infinite vicissitudes of the nations and mankind. Add to them the alterations that all these languages gradually underwent, not only as a consequence of passing time, and still independently of any other circumstance, but also of their ultimately becoming more or less grammatical, the softening of words produced by nascent civilization and by men of letters, according to the varieties of national ear, etc., and also as a result of their being applied, not only in writing, but also in literature. And is there any need now to repeat what has been said so many times about the utmost influence exercised by literature on the modification and formation of languages? Even so, I would point out that the ancient languages have only reached us, not through writing pure and simple, but through literature. On the alterations in meaning to which words are subject, see pp. 1505, etc., and 1501–502.
[1272] And after all that, you will not be surprised if it is so difficult, and very often impossible, to discern their unity of origin in the very different and virtually innumerable languages of the world. Nor will you be surprised if the absolutely original language or languages, or rather that or those first very meager and primitive lists of names, which were the foundation of all languages and still constitute the roots of their words, are submerged in the derivatives, inflections, and compounds which diverged according to the random accidents of the formation of languages, the characters, temperaments, climates, and literatures that shaped those languages, the opinions, customs, and very different circumstances of life which influenced them, the ideas, the dispositions of earth and sky, etc. etc., modified and distorted by the differences in the organs among the different nations, by the ignorance of original speakers, and by the corruption words inevitably suffer, even in the most well-established and perfect languages. You will not be surprised, I repeat, if such primitive roots, though common to all the languages, are, for the most part, hidden from the eyes of the most acute observers, drive the etymologist to despair, and lead him to regard the investigation into the origins of languages as a pipe dream, and with it the study of etymologies and of the analogies between words drawn from every language (a task undertaken, however, by several people, and most recently, from what I have heard, by some Frenchman or other); in short, the original unity of origin and analogy of all languages.1 (Refer all these remarks to what I have said elsewhere [→Z 955, 1022, 1265ff.] on the necessary variety of languages, and conversely those thoughts to these.)
It is just as certain that we very often encounter conformities in outward forms and in the meaning of specific words in the most disparate languages spoken in the very earliest times, by peoples that were very far apart from each other. These words are, in large part, so necessary to life, they express things that are so necessary and, at the same time, so simple, basic, and natural as to demand expression, that for an explanation of these conformities—if we do not wish to attribute them to chance, which is improbable, or to nature, since we are dealing with words whose expression and form are almost entirely arbitrary, [1274] nor can we attribute them to the subsequent relations of the peoples with one another, because that is often at odds with all the known histories, and also because we are dealing with necessary and basic words in all languages—it only remains for us to attribute them to a common origin of these languages and peoples, even though they are now and since the remotest times utterly disparate, distant, and unknown one to the other.
In order, therefore, to discover a common origin of languages and hence of nations (be it a single origin, or a very limited number), to recover as much as we are able of the first language of men, to satisfy the philosophical cravings of that German metaphysician1 (see p. 1134), etc. etc., etymological study is the only means we have. And the only way open to etymology is to profit from the comparative insights provided by the extensive knowledge of many languages, and from profound archaeological and philological, physiological and psychological, etc., insights, and then to take into consideration words from the better known of the most ancient languages (therefore closer to the common origin of languages) and strip them of every inflection, composition, and grammatical derivative, etc. etc., in order to extract the simplest root that it can, and then by means of such comparative insights, etc., take this root back from the very varied alterations in form and in sound that it may have received (even prior to its becoming a root of another word, and in its simplest state, or else afterward) to its original form. Even if the latter cannot be precisely discovered and established, the Etymologist will have done enough, and made what is indeed a highly useful contribution, if he has shown that a specific, demonstrably root word, even though it may be different in the different languages, is however originally a single word, and that among the different forms, meanings, etc., of that root, the original form, meaning, etc., may be found, even though one cannot establish conclusively whether it is one or other of the meanings and forms it has in different languages. We have sought [1275] to indicate how this may be done with the help of philology and archaeology in the case of Latin, which is one of the languages that is oldest, best known, and best suited to such researches, and we have shown how from Latin words one may extract the monosyllabic root. Philosophy helps by teaching us that the first languages must have been predominantly monosyllabic, and composed almost entirely of nouns. We have also shown many accidents in Latin words that were hitherto regarded as essential qualities. This is infinitely detrimental, evidently, to the discovery of ultimate roots and halts the advance of etymological investigations when they are still very far from their goal, and at a point where they ought not to be halted as if they had already reached the ultimate origins and ultimate elements of words. We have, in short, sought to reduce the analysis and decomposition of Latin words to simpler elements. This helps enormously in the knowledge of their origins and roots, just as chemistry made boundless progress when it discovered that the four elements once believed to be original were in fact compounds, and succeeded in discovering substances which, though not themselves wholly elementary and fundamental, are certainly far simpler than those previously known.
[1276] I wish to adduce another example, in addition to those already referred to, in confirmation of the above and to show how much archaeological insights help in the quest for very ancient roots. Silva [forest] is a root in Latin, that is to say, it does not stem from any other known Latin word. Note, however, just how much it has altered from its old and perhaps first form. Almost all etymologists concur that ῾´Υλη is the same as silva.1 Yet how is it that the Latin word has an s and a v more than the Greek? So far as the s is concerned, see what I have noted elsewhere [→Z 1127], see Giulio Pontedera, Antiquitatum Latinarum Graecarumque enarrationes atque emendationes, Letter 2, Padua, Typis Seminarii 1740, p. 18 (the first two letters are worth reading for the light they shed on these archaeological researches regarding the Latin language),2 and it is anyway something already familiar to scholars. In ancient Greek inscriptions, too, the sigma is often to be found in front of words beginning with a vowel, instead of an aspirate. Indeed, this way of writing is often retained in several Greek words (as also in the Latin): e.g., σῦκον [fig] used at first to be pronounced ὗκον or ὖκον with either a hard or a soft breathing, since the Aeolians rendered it ϝὖκον and the Latins ficus. See the Encyclopédie, art. “S.” As for the v, here’s how my argument would go.
The ancient Greek H derived from the Phoenician, Samaritan, and Hebrew Heth, with which it also shares the name ἦτα (since the Greek ταῦ derives from Hebrew thau), along with the shape, etc., was at the beginning simply a sign for an aspirate (see p. 1136, margin), as it still was in Latin, and as it was in the alphabet from which the Greek alphabet came. (See Cellarius, Orthographia, Padua, Comino 1739, p. 40, end and the Encyclopédie méthodique. Grammaire, art. “H,” especially p. 215, and, if you will, Forcellini, under H.) We have seen that the ancient Latin v was simply [1277] the Aeolic digamma, and the latter simply a letter the Aeolians put in place of aspiration, indeed, it was itself a sign of aspiration, and, in short, blood brother of the ancient Greek H. Hence in the very earliest times the word ὕλη used to be pronounced hulh with two aspirates, one at the beginning and the other at the end. (I mean, in short, that the η in ὕλη was not in the beginning a mobile sound, and a pure sign of desinence, but a radical one, which may be deduced from the v the Latins have as a radical sound in this word, that is to say, in silva.) Or else it used to be pronounced hilh, since one cannot fully ascertain what the very earliest pronunciation of the Greek υ was, whether it was u, like the French, as the Greeks pronounced it in their heyday, or i, as modern Greeks pronounce it, as it is pronounced in very many Latin words that were either the daughters or the sisters of Greek words, and as the Germans used to pronounce their ü. What is certain is that the ancient Latins pronounced and wrote those words that in Greek used to be written with Y, sometimes with I and sometimes with u, and hence sometimes also corruptly with o, as from sumnus, somnus [sleep], etc. See Pontedera, loc. cit. on the previous page. Never, however, with y, a Greek character, which “graecorum caussa nominum adscivimus” [“we have adopted on account of Greek nouns”] says Priscian (bk. 1, p. 543, as edited by Putsch), and it is not an ancient character, as Cicero points out, and it used to be pronounced in the Greek fashion, like a French u, to judge by what we find in Martianus Capella. (See Forcellini, the Encyclopédie, and Cellarius, Orthographia, pp. 6, end–7, beginning.) Hence in our case, the ancient marbles and manuscripts, and the scholars, reject the spellings sylva sylvestris, etc., in favor of silva, a corrupt, [1278] and more modern spelling, introduced by barbarian Latin writers, as may be seen in Du Cange. Which furthermore also ser
ves to show the derivation of the Latin silva from, or its cognation with, the Greek ὕλη, there being no other reason why the usage of very ignorant times, which did not think or know anything about etymologies or Greek, should introduce this Greek y into a word that the ancient Latins wrote with an i, a usage retained up until our own times by many who still write sylva and likewise with its derivatives. And perhaps when, according to what Cicero says,1 they began to write and pronounce (that is to say, in exchange for a Gallic u) Pyrrhus and Phryges, etc., instead of the Purrus and Phruges the ancients used to write (see Forcellini, under Y), they also began to write and pronounce sylva. Or at any rate, whenever this happened, its origin and cause lay in the vice of wishing to make writing and pronunciation conform wholly to the foreigners, in the case of words coming from them, a vice which Cicero castigates in the same passage. (A remark highly applicable to the French.) And this, therefore, shows that silva was considered to be the selfsame word as ὕλη, even though writing it as sylva is corrupt. It is again the case with the Greeks in their heyday that words with a υ, when they undergo the changes customary in Greek words, often change the υ into ι, as from δύο [two] δὶς [twice] is made, and likewise with compounds (such as διπλοῦς [double], διττὸς [double], δίστομος [double-mouthed], διφυὴς [having a double nature], etc.), it is always δι-.
Reverting to my argument, both today and for a long time now, this same Greek letter y, introduced into the Latin alphabet solely to represent the Greek υ, and to express the sound of the French u, [1279] is invariably pronounced in that alphabet, as in the language, as a pure i. So too in Spanish and in French, when it is not turned into an i in the writing, too, as it always is in our language. And note that in these two languages the y is also pronounced i in words or proper names, etc., which were not derived from Latin, or which in Latin did not have this letter, or even had i instead. And y and i are continuously interchangeable in the written forms of Spanish and French, especially in those which are not wholly modern, since today the spelling is more fixed. (The French write Sylvain, pronouncing it Silvain. See too the Spanish Dictionary under Syl.)1 Note once again that the French retain the Gallic u, and also pronounce y as i. Which suggests that this latter Greco-Latin letter wholly and invariably lost its original sound, and turned into i, like the υ for the Greeks. And it is natural for there to be a reciprocal affinity between i and u, since they are the thinnest of our vowels. See p. 2152, end. In point of fact the sound of the French or Lombard u (Forcellini calls it Bergamasque) has something of the i and the u alike. And those same Greeks who pronounced their υ as the French do u, regarded it as an i rather than as a u, I mean as a variant or inflection, etc., of i. Since in their alphabet they called it ὑψιλὸν (as we likewise call it in Greek fashion ipsilon), that is to say, weak υ. Now this adding on of weak is simply designed to make a distinction, just as ε is called ἐψιλὸν so as to distinguish it from ἦτα. But in their alphabets, the Greeks do not have another u from which they have to distinguish this υ, although they do indeed have another i, that is, ἰῶτα.
Zibaldone Page 111