Billeder på siden

extent. The process is well exemplified by the series Ausones, itself evidently formed by a familiar suffix, on (en) (i.e. aus + on) Auruncus, Aurunculus (cf. sermunculus) Aurunculeius (cf. Pompeius). If this word had not been appropriated to an individual or family we might have gone on to † Aurunculeitas (cf. Appietas, actually made by Cicero) or †Aurunculeare and from that to †Aurunculeatura or farther still unless the word should break down under its own weight.

Another principle is that in this continued process two or more of the elements successively added become fused together so as to be regarded as a single suffix and used as such. For instance, in the series mentioned, unculus comes to be felt as an integral element and so is applied as a whole to words where the intervening steps do not appear and probably never existed. Thus we have avunculus, without any avo (-onis), or avuncus. By comparing several series of words, however, we can in almost all cases recognize the steps of the process.

These two principles of stem formation have been followed in the Latin language more, perhaps, than in any other, so that the Latin vocabulary is particularly rich in such long and highly developed words, whereby the shorter have been in great measure superseded.

Another principle is that words in the process of development tend to become specialized in particular meanings. In their origin adjectival in sense, i.e. expressing qualities either active or passive or sometimes both indifferently, they become participles, nouns of agency, names of instruments, or even, more exactly, names of persons, places, or of any idea that seeks expression in human speech.

A fourth principle, not so freely recognized, but to my mind equally certain, is that scientifically no derivative is strictly (i.e. originally) made either from a verb or noun as such. Derivation evidently goes far back of any such distinction as verb and noun. The elements used were neither verbs nor nouns, because they were both at once, and in this state of language the type of derivation was fixed. The later derivatives, consciously made, come from parts of words abstracted as stems and treated in the old manner just as if inflexion had never existed. Any form that seemed like a root or a stem could be conceived as a suitable element for further formation on the fixed pattern.

Thus the word laudator is not in a strict sense derived from laudare. The tor type must have become fixed long before there was

any verb or noun like laus or laudo. Such forms as actor, genitor, were early developed in the language and had become attached to agere and gignere as nouns of agency for those verbs. So on the same pattern were produced laudator, auditor, and the like. All this depends on the principle that composition and stem formation preceded in idea and type any inflexion or distinction of parts of speech whatsoever.

Our principles then are:

First. Stem formation by successive addition of suffixes. Second. The fusing together of two or more of these suffixes so as to make a new available one.

Third. The specialization of the meanings of the words at any stage of their development.

Fourth. Derivation proceeds by stems and antedates inflexion and parts of speech.

In view of these principles, when we find the long words which are so characteristic of Latin, the natural presumption is that (apart from obvious composition) the words have been built up by continuous further formation by means of the living elements existing in the language, and unless some controlling reason appears to the contrary, this presumption is to be taken as true. As the suffixes are for the most part of pronominal origin we must, in analyzing a word, take off successively from the end the recognizable suffixes and discover the stems or the various steps through which the word has passed in its further formation.

Now it is noticeable in Latin that among the numerous derivatives there are a number of sets or series of words, in which each word has the same final letters with different letters in the body of the word, but with only slight differences in meaning. We have, for instance, figura (the only one of its kind) alongside of pictura (one of a large number). If we proceed by the method above indicated, we find in one case a root fig + stem-suffix u + stem-suffix ra, in the other pig + tu + ra. It seems obvious that we have here two differently formed stems continued by the same suffix ra.

In a pair of somewhat similar formation, maturus and Matuta, the same stem is continued by means of different suffixes. Compare this

+matu- with mane, and we see ma + tu and ma + ne. So we may conclude that the much-used tura is a compound suffix formed of tu + ra and is really the feminine of tu + ro.1

Again we have rationalis, rationabilis, and ratiocinabiliter (implying a ↑ ratiocinabilis). We instantly recognize ration + ā + lis, ration + a + bilis, and ratiocinā + bilis. This process, which is a well-known one, ought to be carried still further, so that the ultimate analysis of the last form should be, on the same principle, ra + ti + on + co + no + a (representing the formative elements of a verb stem, treated, however, according to old patterns as a productive stem in conformity with the fourth principle above) + bo+lis. We notice in the process that ra + ti may be bracketed, as in mens, mentis; that ti + on may be bracketed, as in mentio, mentionis; that co+no may be bracketed, as in lenocinium; and that bo + lis are fused in the same manner. These again are often fused with ā, as we ultimately get the practical suffix in our bearable. So the steps are † ratis, ratio, † ratiocinus, ratiocinor, ratiocinabilis (as implied in the adverb). It is to be noticed that in any single word we can rarely be sure of the chronology, so as to know whether the fusion of the suffixes came before or after the formation of the particular word, but by comparison we may always be sure of the type, and may confidently by means of daggers give the typical intermediate steps. This to my mind is the only proper way of analyzing words so as to give certain conclusions. Let us apply it to less obvious cases.

Some most difficult series are those in

[blocks in formation]

As we see, the letters vary in the middle, but the last elements are the same. The words in each of these series have nearly the same meaning, and in view of the principles laid down we may assume on the face of the matter that the varying letters come from different stems, i.e. from the use of different suffixes at some stage of the pro

1 The length of the u is only incidental, and need not be considered here.

gressive development of the words. They differ from each other just as the words

ferax felix ferox fiducia




which have the same x or cus ending, but are formed from different verb stems.


Let us see, then, whether we have any warrant for assuming the successive stems and the successive suffixes in these groups, as we found them in Auso, Auruncus, † Aurunculus Aurunculeius.

We take in the first series, say, fragilis, nobilis, versatilis. According to our first principle, these have been made by continuous further formation. As elements we have a root frag; we have a common suffix o/a. These have given us fragus (cf. silvifragus). We have also a suffix li. This gives fragilis, "breakable," a type which is well represented in the language (cf. agilis, habilis, docilis, bibilis). We may notice in passing that in accordance with our third principle the type has become specialized in the sense of passive capability, though no such sense seems inherent in the suffix. The words are adjectival in sense originally (cf. herbilis from herba). The type, however, has, as so often happens in Latin, been supplanted by other longer forms. Suppose we proceed in the same manner with the others. We have a stem versato (versatus), which seems to be treated exactly as the simpler form fragus. The result is naturally versatilis. This also remained as a type (cf. coctilis, fissilis, flexilis, volatilis), so that we finally have fluviatilis, "belonging to a river.". The meaning of this last word is a clear indication that the original force of the compound ending tilis was an adjective one, not exactly defined in any one function, as it generally became later. The change in meaning may be illustrated by 'a catchy melody,' i.e. one easily caught. I can see no reason why we should not proceed in the same manner with nobilis. We have a root (or stem) no, as in notus. But here we haven't any † nobus. Still we do have morbus, turba, manubiae (implying a manubus), tubus (cf. tumeo), tribus (tres), dubo (are), dubium, addubanum, and herba (whence herbilis). We have also many forms which, treated on the same principle, show a b element as a component part. Such are ber, bris, bre; ber, bra, brum; bulus, -a, -um; bundus,

-a, -um; and we may probably reckon trabes, trabea, and plebes. We shall also find the same phenomena in the other series hereafter to be discussed. Why may we not, then, assume a † nobus (like morbus, herba, † manubus) further formed by lis, as in fact herbilis is formed?

The second series has for example celeris, mediocris, celebris, equestris. For variety we may also give alebris, anclabris, October, tuber. Proceeding as before, we have a stem cele obviously akin to cello.1 Added to this we have ri (a well-known suffix, like li in the first series) making celeris often, phonetically, celer.

In the second word we have a stem medio and the common suffix ko/a, which would make † mediocus. This form is fortunately proved by medioximus, especially medioxume, an odd superlative of mediocriter (medioc+timus, cf. oxime). To this we may confidently add ri (cf. li in the first series) making mediocris. For a parallel to † mediocus we may cite alica, "spelt" from alo.

In alebris we have a stem akin to alo, precisely as we have cele in celebris and celeris from cello. The natural presumption would be that this stem was further formed with a suffix (see first principle), just as we have alica with the ka formation. If alicris had happened to result we should have seen the connection at once. But the bo/a formation was so meagerly retained that we are driven to conjecture. Still we have all the forms mentioned before,—morbus, turba, herba, and the analogies of the other suffixes. Particularly we may compare manubiae by the side of manubrium. It is to be noticed that the force of these comparisons grows in geometrical proportion with every additional analogy. Why may we not suppose an † aleba (or bus), like morbus and turba? This is now ready for a further formation with ri again, giving the form alebris, as we have it. (We may here compare alibilis, its synonym that later supplanted it.) In this way bris becomes a suffix to be used as in muliebris, anclabris, October, all with a general adjective meaning, and tuber specialized as a noun. These show that there was no definitive idea attached as yet to the termination.

For the tris formation, as in palustris, equestris, we have no direct

1 The relation of this e to the o and i suffixes is not clear, but the interchange is a common one.

« ForrigeFortsæt »