Varro: DE LINGVA LATINA

In verborum declinationibus disciplina loquendi dissimilitudinem an similitudinem sequi deberet, multi quaesierunt. Cum ab his ratio quae ab similitudine oriretur vocaretur analogia, reliqua pars appellaretur anomalia: de qua re primo libro quae dicerentur cur dissimilitudinem ducem haberi oporteret, dixi, secundo contra quae dicerentur, cur potius similitudinem conveniret praeponi: quarum rerum quod nec fundamenta, ut debuit, posita ab ullo neque ordo ac natura, ut res postulat, explicita, ipse eius rei formam exponam.

In the declensions of words, whether the discipline of speaking ought to follow dissimilarity or similarity, many have inquired. Since by these the reason which would arise from similarity is called analogy, the remaining part is called anomaly: on which matter, in the first book I said what would be said as to why dissimilarity ought to be held as the guide, in the second, what would be said in reply, as to why it would be more fitting to prefer similarity: of which things, because neither the foundations, as was due, have been laid by anyone, nor the order and nature, as the matter demands, unfolded, I myself will set forth the form of this matter.

Dicam de quattuor rebus, quae continent declinationes verborum: quid sit simile ac dissimile, quid ratio quam appellant logon, quid pro portione quod dicunt ana logon, quid consuetudo; quae explicatae declarabunt analogiam et anomaliam, unde sit, quid sit, cuius modi sit.

I will speak about four things, which contain the declensions of words: what the similar and the dissimilar are, what the rationale which they call logon is, what proportion, which they call ana logon, is, what custom is; which, once explained, will declare analogy and anomaly—whence it is, what it is, of what sort it is.

II.

De similitudine et dissimilitudine ideo primum dicendum, quod ea res est fundamentum omnium declinationum ac continet rationem verborum. Simile est quod res plerasque habere videtur easdem quas illud cuiusque simile: dissimile est quod videtur esse contrarium huius. Minimum ex duobus constat omne simile, item dissimile, quod nihil potest esse simile, quin alicuius sit simile, item nihil dicitur dissimile, quin addatur quoius sit dissimile.

On similitude and dissimilitude it must therefore be said first, because that matter is the foundation of all declensions and contains the rationale of words. A similitude is what seems to have most things the same as that to which it is similar; a dissimilitude is what seems to be the contrary of this. Every similitude consists, at the minimum, of two, likewise every dissimilitude, because nothing can be similar without being similar to something, and likewise nothing is said dissimilar without adding whose it is dissimilar.

Sic dicitur similis homo homini, equus equo, et dissimilis homo equo: nam similis est homo homini ideo, quod easdem figuras membrorum habent, quae eos dividunt ab reliquorum animalium specie. In ipsis hominibus simili de causa vir viro similior quam vir mulieri, quod plures habent easdem partis; et sic senior seni similior quam puero. Eo porro similiores sunt qui facie quoque paene eadem, habitu corporis, filo: itaque qui plura habent eadem, dicuntur similiores; qui proxume accedunt ad id, ut omnia habeant eadem, vocantur gemini, simillimi.

Thus a man is said to be similar to a man, a horse to a horse, and a man dissimilar to a horse: for a man is similar to a man for this reason, that they have the same figures of the members, which separate them from the species of the remaining animals. Among men themselves, for a like cause a man is more similar to a man than a man to a woman, because they have more of the same parts; and so an elder is more similar to an elder than to a boy. Moreover, those are more similar who also are almost the same in face, in the habit of the body, in hair; and thus those who have more things the same are called more similar; those who come most proximate to this, that they have all things the same, are called twins, most similar.

Sunt qui tris naturas rerum putent esse, simile, dissimile, neutrum, quod alias vocant non simile, alias non dissimile (sed quamvis tria sint simile dissimile neutrum, tamen potest dividi etiam in duas partes sic, quodcumque conferas aut simile esse aut non esse); simile esse et dissimile, si uideatur esse ut dixi, neutrum, si in neutram partem praeponderet, ut si duae res quae conferuntur vicenas habent partes et in his denas habeant easdem, denas alias ad similitudinem et dissimilitudinem aeque animadvertendas: hanc naturam plerique subiciunt sub dissimilitudinis nomen.

There are those who think there are three natures of things: similar, dissimilar, and neutral—which some call “not similar,” others “not dissimilar” (but although similar, dissimilar, neutral are three, nevertheless it can also be divided into two parts thus: whatever you compare is either to be similar or not to be); “similar” and “dissimilar,” if it appears to be as I have said; “neutral,” if it preponderates to neither side, as when two things that are being compared have twenty parts apiece and among these have ten the same, ten others equally to be observed for similarity and dissimilarity: most people subject this nature under the name of dissimilarity.

Quare quoniam fit ut potius de vocabulo quam de re controversia esse videatur, illud est potius advertendum, quom simile quid esse dicitur, cui parti simile dicatur esse (in hoc enim solet esse error), quod potest fieri ut homo homini similis non sit, ut multas partis habeat similis et ideo dici possit similis habere oculos, manus, pedes, sic alias res separatim et una plures.

Wherefore, since it comes about that the controversy seems to be rather about the vocable than about the thing, this rather is to be adverted to: when something is said to be similar, to which part it is said to be similar (for the error is wont to be in this), because it can happen that a man is not similar to a man, in that he has many parts similar and therefore can be said to have similar eyes, hands, feet, and so other things separately and several together.

Itaque quod diligenter videndum est in verbis, quas partis et quot modis oporteat similis habere quae similitudinem habere dicuntur, ut infra apparebit, is locus maxime lubricus est. Quid enim similius potest videri indiligenti quam duo verba haec suis et suis? Quae non sunt, quod alterum significat suere, alterum suem.

Therefore, in words, the matter which must be looked at diligently—namely, which parts and in how many modes those things which are said to have similitude ought to have like parts, as will appear below—is a most slippery topic. For what could seem more similar to the careless than these two words, suis and suis? Which are not, since the one signifies “to sew,” the other “a sow.”

Itaque similia vocibus esse ac syllabis confitemur, dissimilia esse partibus orationis videmus, quod alterum habet tempora, alterum casus, quae duae res vel maxime discernunt analogias.

Therefore we confess that they are similar in voices and syllables; we see that they are dissimilar in the parts of speech, because the one has tenses, the other cases—two things which most especially distinguish analogies.

Item propinquiora genere inter se verba similem saepe pariunt errorem, ut in hoc, quod nemus, et lepus videtur esse simile, quom utrumque habeat eundem casum rectum; sed non est simile, quod eis certae similitudines opus sunt, in quo est ut in genere nominum sint eodem, quod in his non est: nam in virili genere est lepus, ex neutro nemus; dicitur enim hic lepus et hoc nemus. Si eiusdem generis essent, utrique praeponeretur idem ac diceretur aut hic lepus et hic nemus aut hoc nemus, hoc lepus.

Likewise, words nearer in genus to one another often beget a similar error, as in this, that nemus and lepus seem to be similar, since each has the same nominative (the “straight”) case; but they are not similar, because fixed likenesses are required for them—among which is that they be in the same gender of nouns—which is not the case with these: for lepus is of the masculine gender, whereas nemus is of the neuter; for one says “hic lepus” and “hoc nemus.” If they were of the same gender, the same would be set before each, and it would be said either “hic lepus” and “hic nemus,” or “hoc nemus, hoc lepus.”

Quare quae et cuius modi sunt genera similitudinum ad hanc rem, perspiciendum ei qui declinationes verborum proportione sintne quaeret. Quem locum, quod est difficilis, qui de his rebus scripserunt aut vitaverunt aut inceperunt neque adsequi potuerunt.

Wherefore, what and of what sort the kinds of similitudes are for this matter must be scrutinized by the one who will inquire whether the declensions of words are by proportion or not. Which topic, because it is difficult, those who have written about these matters have either avoided, or, having begun, have not been able to attain.

Itaque in eo dissensio neque ea unius modi apparet: nam alii de omnibus universis discriminibus posuerunt numerum, ut Dionysius Sidonius, qui scripsit ea esse septuaginta unum, alii partis eius quae habet casus, cuius eidem hic cum dicat esse discrimina quadraginta septem, Aristocles rettulit in litteras XIIII, Parmeniscus VIII, sic alii pauciora aut plura.

Accordingly, in this there appears dissension, and not of one kind: for some have set a number for all the universal distinctions, as Dionysius Sidonius, who wrote that these are 71; others for that part of it which has cases, of which this same man, when he says that there are 47 distinctions, Aristocles committed to writing 14, Parmeniscus 8, thus others fewer or more.

Quarum similitudinum si esset origo recte capta et inde orsa ratio, minus erraretur in declinationibus verborum. Quarum ego principia prima duum generum sola arbitror esse, ad quae similitudines exigi oporteat: e quis unum positum in verborum materia, alterum ut in materiae figura, quae ex declinatione fit.

Of these similitudes, if the origin were rightly seized and the method begun from there, there would be less error in the declinations of words. Of which I judge the first principles to be of two kinds only, to which the similitudes ought to be referred: of which one is set in the matter of words, the other as in the figure of the matter, which is made from declension.

Nam debet esse unum, ut verbum verbo, unde declinetur, sit simile; alterum, ut e verbo in verbum declinatio, ad quam conferetur, eiusdem modi sit: alias enim ab similibus verbis similiter declinantur, ut ab erus ferus, ero fero, alias dissimiliter erus ferus, eri ferum. Cum utrumque et verbum verbo erit simile et declinatio declinationi, tum denique dicam esse simile ac duplicem et perfectam similitudinem habere, id quod postulat analogia.

For there ought to be one condition: that the word be similar to the word from which it is declined; a second: that the declension from word into word, to which it will be compared, be of the same mode; for sometimes from similar words they are declined similarly, as from erus, ferus, ero, fero; sometimes dissimilarly, erus, ferus, eri, ferum. When both are present—both word to word will be similar and declension to declension—then at last I shall say that it is similar and has a double and perfect similitude, that which analogy demands.

Sed ne astutius videar posuisse duo genera esse similitudinum sola, cum utriusque inferiores species sint plures, si de his reticuero, ut mihi relinquam latebras, repetam ab origine similitudinum quae in conferendis verbis et inclinandis sequendae aut vitandae sint.

But lest I seem slyly to have posited that there are only two genera of similitudes, since the inferior species of each are more numerous, if I keep silent about these, so that I may leave hiding-places to myself, I will go back to the origin of the similitudes which, in comparing words and in inflecting, are to be followed or avoided.

Prima divisio in oratione, quod alia verba nusquam declinantur, ut haec vix mox, alia declinantur, ut ab lima limae, a fero ferebam, et cum nisi in his verbis quae declinantur non possit esse analogia, qui dicit simile esse mox et nox errat, quod non est eiusdem generis utrumque verbum, cum nox succedere debeat sub casuum rationem, mox neque debeat neque possit.

The first division in discourse is this: that some words are nowhere declined, as these “vix,” “mox,” others are declined, as from “lima” “limae,” from “fero” “ferebam”; and since analogy cannot exist except in those words which are declined, he who says that “mox” and “nox” are similar errs, because the two words are not of the same genus, since “nox” ought to come under the rule of cases, while “mox” neither ought nor can.

Secunda divisio est de his verbis quae declinari possunt, quod alia sunt a voluntate, alia a natura. Voluntatem appello, cum unus quivis a nomine aliae rei imponit nomen, ut Romulus Romae; naturam dico, cum universi acceptum nomen ab eo qui imposuit non requirimus quemadmodum is velit declinari, sed ipsi declinamus, ut huius Romae, hanc Romam, hac Roma. De his duabus partibus voluntaria declinatio refertur ad consuetudinem, naturalis ad rationem.

The second division is about those words which can be declined, namely that some are by will, others by nature. I call it will, when some individual imposes a name from the name of another thing, as Romulus to Rome; I call it nature, when, the name having been received by all, we do not require from the one who imposed it in what manner he wishes it to be declined, but we ourselves decline it, as of this Rome, this Rome, with this Rome. Of these two parts, voluntary declension is referred to custom, natural to reason.

Quare proinde ac simile conferre non oportet ac dicere, ut sit ab Roma Romanus, sic ex Capua dici oportere Capuanus, quod in consuetudine vehementer natat, quod declinantes imperite rebus nomina imponunt, a quibus cum accepit consuetudo, turbulenta necesse est dicere. Itaque neque Aristarchei neque alii in analogiis defendendam eius susceperunt causam, sed, ut dixi, hoc genere declinatio in communi consuetudine verborum aegrotat, quod oritur e populo multiplici et imperito: itaque in hoc genere in loquendo magis anomalia quam analogia.

Wherefore it is not proper to bring together as similar and to say: just as from Rome there is Romanus, so from Capua Capuanus ought to be said; for this floats violently in consuetude, because those declining words ignorantly impose names upon things, from which, when consuetude has received them, it is needful to call it turbulent. And so neither the Aristarchians nor others within analogies undertook the case that it be defended; but, as I said, in this kind declension in the common consuetude of words is diseased, because it arises from a populace manifold and unskilled: and so in this kind, in speaking, there is more anomaly than analogy.

Tertia divisio est: quae verba declinata natura; ea dividuntur in partis quattuor: in unam quae habet casus neque tempora, ut docilis et facilis; in alteram quae tempora neque casus, ut docet facit; in tertiam quae utraque, ut docens faciens; in quartam quae neutra, ut docte et facete. Ex hac divisione singulis partibus tres reliquae dissimiles. Quare nisi in sua parte inter se collata erunt verba, si conveniunt, non erit ita simile, ut debeat facere idem.

The third division is: which words are declined by nature; these are divided into four parts: into one which has cases and not tenses, as docilis and facilis; into another which has tenses and not cases, as docet, facit; into a third which has both, as docens, faciens; into a fourth which has neither, as docte and facete. From this division, for each single part the remaining three are dissimilar. Therefore, unless words, when compared with one another within their own part, do agree, it will not be so similar that it ought to do the same.

Unius cuiusque partis quoniam species plures, de singulis dicam. Prima pars casualis dividitur in partis duas, in nominatus scilicet et articulos, quod aeque finitum et infinitum est ut hic et quis; de his generibus duobus utrum sumpseris, cum reliquo non conferendum, quod inter se dissimiles habent analogias.

Since each single part has several species, I will speak about the individual ones. The first, the case-part, is divided into two parts, namely into Nominate and Articles, which are equally finite and infinite, as hic and quis; of these two genera, whichever you take, it is not to be compared with the remainder, because they have analogies dissimilar among themselves.

In articulis vix adumbrata est analogia et magis rerum quam vocum; in nominatibus magis expressa ac plus etiam in vocibus ac syllabarum similitudinibus quam in rebus suam optinet rationem. Etiam illud accedit ut in articulis habere analogias ostendere sit difficile, quod singula sint verba, hic contra facile, quod magna sit copia similium nominatuum. Quare non tam hanc partem ab illa dividendum quam illud videndum, ut satis sit verecundiae etiam illam in eandem arenam vocare pugnatum.

In the articles the analogy is scarcely adumbrated, and is more of things than of words; in the nominals it is more expressed, and it obtains its own reason more in words and in similarities of syllables than in things. There is also this in addition: that in the articles it is difficult to show that they have analogies, because they are single words; here, on the contrary, it is easy, because there is a great abundance of similar nouns. Wherefore, this part is not so much to be divided from that, as this must be considered: that it is enough for modesty to call even that one into the same arena to fight.

Ut in articulis duae partes, finitae et infinitae, sic in nominatibus duae, vocabulum et nomen: non enim idem oppidum et Roma, cum oppidum sit vocabulum, Roma nomen, quorum discrimen in his reddendis rationibus alii discernunt, alii non; nos sicubi opus fuerit, quid sit et cur, ascribemus.

As in the articles there are two parts, the finite and the infinite, so in nouns there are two, vocable and name: for a town and Rome are not the same, since town is a vocable, Rome a name—the discrimination of which, in rendering these accounts, some discern, others do not; we, wherever there shall be need, will ascribe what it is and why.

Nominatui ut similis sit nominatus, habere debet ut sit eodem genere, specie eadem, sic casu, exitu eodem: specie, ut si nomen est quod conferas, cum quo conferas sit nomen; genere, ut non solum unum sed utrumque sit virile; casu, ut si alterum sit dandi, item alterum sit dandi; exitu, ut quas unum habeat extremas litteras, easdem alterum habeat.

For a noun to be similar to the noun to which it is compared, it ought to have this: that it be of the same genus, the same species, likewise the same case, the same ending: in species, such that if what you bring to compare is a name, that with which you compare be a name; in genus, that not only the one but both be masculine; in case, that if the one be of giving, likewise the other be of giving; in ending, that the letters which the one has as final, the other have the same.

Ad hunc quadruplicem fontem ordines deriguntur bini, uni transversi, alteri derecti, ut in tabula solet in qua latrunculis ludunt. Transversi sunt qui ab recto casu obliqui declinantur, ut albus albi albo; derecti sunt qui ab recto casu in rectos declinantur, ut albus alba album; utrique sunt partibus senis. Transversorum ordinum partes appellantur casus, derectorum genera, utrisque inter se implicatis forma.

To this fourfold spring the ranks are directed in pairs, one transverse, the other straight, as on a board on which they play Latrunculi. Transverse are those which from the direct case are declined oblique, as albus albi albo; straight are those which from the direct case are declined into direct cases, as albus alba album; each has six parts. The parts of the transverse orders are called cases, those of the straight, genders, and in both, form is intertwined among them.

Dicam prius de transversis. Casuum vocabula alius alio modo appellavit; nos dicemus, qui nominandi causa dicitur, nominandi vel nominativum ***

I will speak first about the transverse ones. As for the vocabula of the cases, one has named them in one way, another in another; we will say that that which is said for the sake of naming is “of naming” or “nominative” ***

*** dicuntur unae scopae, non dicitur una scopa: alia enim natura, quod priora simplicibus, posteriora in coniunctis rebus vocabula ponuntur, sic bigae, sic quadrigae a coniunctu dictae. Itaque non dicitur, ut haec una lata et alba, sic una biga, sed unae bigae, neque dicitur ut hae duae latae, albae, sic hae duae bigae et quadrigae, sed hae binae bigae et quadrigae.

*** they say “one brooms” (unae scopae), one does not say “one broom” (una scopa): a different nature, since former words are put for simple things, latter for conjoined things; thus bigae, thus quadrigae are said from a conjunction. And so it is not said, as “this one broad and white” (haec una lata et alba), thus “one biga,” but “one bigae” (unae bigae); nor is it said, as “these two broad, white,” thus “these two bigae and quadrigae,” but “these two sets (in pairs) of bigae and quadrigae” (hae binae bigae et quadrigae).

Item figura verbi qualis sit refert, quod in figura vocis alias commutatio fit in primo verbo suit modo suit, alias in medio, ut curso cursito, alias in extremo, ut doceo docui, alias communis, ut lego legi. Refert igitur ex quibus litteris quodque verbum constet, maxime extrema, quod ea in plerisque commutatur.

Likewise, it matters what sort the figure of the verb is, because in the figure of the voice sometimes a commutation happens in the first part of the word, as suit just now suit, sometimes in the middle, as curso cursito, sometimes at the end, as doceo docui, sometimes a common one, as lego legi. Therefore it is relevant from which letters each verb consists, especially the final ones, because these in very many are commuted.

Quare in his quoque partibus similitudines ab aliis male, ab aliis bene quod solent sumi in casibus conferendis, recte an perperam videndum; sed ubicumque commoventur litterae, non solum eae sunt animadvertendae, sed etiam quae proxumae sunt neque moventur: haec enim vicinitas aliquantum potest in verborum declinationibus.

Wherefore in these parts also, the similarities—by some badly, by others well—which are wont to be taken in comparing cases, must be considered whether rightly or wrongly; but wherever letters are set in motion, not only those are to be taken notice of, but also those which are proximate and are not moved: for this vicinity can do somewhat in the declensions of words.

In quis figuris non ea similia dicemus quae similis res significant, sed quae ea forma sint, ut eius modi res similis ex instituto significare plerumque soleant, ut tunicam virilem et muliebrem dicimus non eam quam habet vir aut mulier, sed quam habere ex instituto debet: potest enim muliebrem vir, virilem mulier habere, ut in scaena ab actoribus haberi videmus, sed eam dicimus muliebrem, quae de eo genere est quo indutui mulieres ut uterentur est institutum. Ut actor stolam muliebrem sic Perpenna et Caecina et Spurinna figura muliebria dicuntur habere nomina, non mulierum.

In which figures we will not call similar those that signify similar things, but those which are of such a form that things of that sort are for the most part accustomed by convention to signify as similar, as we call a tunic virile and muliebral not the one which a man or a woman has, but the one which by institution he or she ought to have: for a womanly one a man can have, a manly one a woman can have, as we see had on the stage by actors; but we call that womanly which is of that kind for which it has been instituted that women should use it for clothing. As an actor has a woman’s stola, so Perpenna and Caecina and Spurinna are said to have feminine figures in their names, not women’s.

Flexurae quoque similitudo videnda ideo quod alia verba quam viam habeant ex ipsis verbis, unde declinantur, apparet, ut quemadmodum oporteat uti praetor consul, praetori consuli; alia ex transitu intelleguntur, ut socer macer, quod alterum fit socerum, alterum macrum, quorum utrumque in reliquis a transitu suam viam sequitur et in singularibus et in multitudinis declinationibus. Hoc fit ideo quod naturarum genera sunt duo quae inter se conferri possunt, unum quod per se videri potest, ut homo et equus, alterum sine assumpta aliqua re extrinsecus perspici non possit, ut eques et equiso uterque enim dicitur ab equo.

The likeness of inflections too must be considered, for this reason: that for some words, which way they take appears from the very words from which they are declined, as how it is proper to use praetor consul, praetori consuli; others are understood from the transition, as socer macer, because the one becomes socerum, the other macrum, each of which in the remaining cases follows its own way from the transition both in the singulars and in the declensions of the multitude. This happens because there are two kinds of natures that can be compared with one another: one which can be seen by itself, as man and horse, the other which cannot be perceived without some thing assumed from outside, as horseman and equerry—for each is said from horse.

Quare hominem homini similem esse aut non esse, si contuleris, ex ipsis hominibus animadversis scies; at duo inter se similiterne sint longiores quam sint eorum fratres, dicere non possis, si illos breviores cum quibus conferuntur quam longi sint ignores; sic latiorum atque altiorum, item cetera eiusdem generis sine assumpto extrinsecus aliquo perspici similitudines non possunt. Sic igitur quidam casus quod ex hoc genere sunt, non facile est dicere similis esse, si eorum singulorum solum animadvertas voces, nisi assumpseris alterum, quo flectitur in transeundo vox.

Wherefore whether a man is like a man or not, if you compare, you will know from the men themselves once observed; but whether two are, between themselves, similarly longer than their brothers are, you cannot say, if you do not know how long those shorter ones are with whom they are compared; thus the likenesses of broader and taller things, and likewise the rest of the same kind, cannot be perceived without something assumed from outside. Thus, therefore, for certain cases which are of this kind, it is not easy to say that they are similar, if you observe only the single forms of each, unless you have assumed the other, by which the word-form is inflected in passing.

Quod ad nominatuom similitudines animadvertendas arbitratus sum satis esse tangere, haec sunt. Relinquitur de articulis, in quibus quaedam eadem, quaedam alia. De quinque enim generibus duo prima habent eadem, quod sunt et virilia et muliebria et neutra, et quod alia sunt ut significent unum, alia ut plura, et de casibus quod habent quinos: nam vocandi voce notatus non est.

As to the likenesses of nouns to be observed, these are what I have judged it sufficient to touch upon. There remains the matter of the articles, in which some things are the same, some different. For of the five genera the first two have the same features, namely that there are masculine and feminine and neuter, and that some are such as to signify one, others to signify more, and, concerning the cases, that they have five: for the voice of calling is not marked.

Proprium illud habent, quod partim sunt finita, et hic haec, partim infinita, ut quis et quae, quorum quod adumbrata et tenuis analogia, in hoc libro plura dicere non necesse est.

They have this proper characteristic, that partly they are definite, namely hic and haec, and partly indefinite, such as quis and quae; concerning which, since the analogy is adumbrated and tenuous, it is not necessary to say more in this book.

Secundum genus quae verba tempora habent neque casus, sed habent personas. Eorum declinatuum species sunt sex: una quae dicitur temporalis, ut legebam gemebam, lego, gemo; altera personarum, ut sero meto, seris metis; tertia rogandi, ut scribone legone, scribisne legisne. Quarta respondendi, ut fingo pingo, fingis pingis; quinta optandi, ut dicerem facerem, dicam faciam; sexta imperandi, ut cape rape, capito rapito.

The second kind is those words which have tenses and not cases, but have persons. The species of their inflections are six: one which is called temporal, as legebam gemebam, lego, gemo; a second, of persons, as sero meto, seris metis; a third, of asking, as scribone legone, scribisne legisne. A fourth, of answering, as fingo pingo, fingis pingis; a fifth, of wishing, as dicerem facerem, dicam faciam; a sixth, of commanding, as cape rape, capito rapito.

Item sunt declinatuum species quattuor quae tempora habent sine personis: in rogando, ut foditurne sereturne, et fodieturne sereturne. Ab respondendi specie eaedem figurae fiunt extremis syllabis demptis; optandi species, ut vivatur ametur, viveretur amaretur. Imperandi declinatus sintne habet dubitationem et eorum sitne haec ratio: paretur pugnetur, parator pugnator.

Likewise there are four species of declensions which have tenses without persons: in asking, as “foditurne sereturne,” and “fodieturne sereturne.” From the species of answering the same figures are made with the final syllables removed; the species of wishing, as “vivatur ametur, viveretur amaretur.” The declension of commanding has a doubt whether there be such forms and whether this be their rationale: “paretur pugnetur, parator pugnator.”

Accedunt ad has species a copulis divisionum quadrinis: ab infecti et perfecti, ut emo edo, emi edi; ab semel et saepius, ut scribo lego, scriptito lectito; a faciendi et patiendi, ut uro ungo, uror ungor; a singulari et multitudinis, ut laudo culpo, laudamus culpamus. Huius generis verborum cuius species exposui quam late quidque pateat et cuius modi efficiat figuras, in libris qui de formulis verborum erunt diligentius expedietur.

Added to these species are fourfold couplings of divisions: from the imperfect and the perfect, as emo edo, emi edi; from once and more-often, as scribo lego, scriptito lectito; from doing and suffering, as uro ungo, uror ungor; from the singular and of the multitude, as laudo culpo, laudamus culpamus. For this kind of verbs, whose species I have set forth, how widely each extends and of what sort it produces figures will be more carefully set forth in the books which will be about the formulas of verbs.

Tertii generis, quae declinantur cum temporibus ac casibus ac vocantur a multis ideo participalia, sunt hoc genere ***

Of the third genus, those which are declined with tenses and cases and are for that reason by many called participials, there are in this genus ***

*** quemadmodum declinemus, quaerimus casus eius, etiamsi siqui finxit poeta aliquod vocabulum et ab eo casum ipse aliquem perperam declinavit, potius eum reprehendimus quam sequimur. Igitur ratio quam dico utrubique, et in his verbis quae imponuntur et in his quae declinantur, neque non etiam tertia illa, quae ex utroque miscetur genere.

*** how we should decline, we inquire its cases; and even if some poet has fashioned some word and from it himself has wrongly declined some case, we rather reprehend him than follow him. Therefore the rationale which I mention holds in both places, both in those words which are imposed and in those which are declined, and likewise also that third, which is mixed from both kinds.

Quarum una quaeque ratio collata cum altera aut similis aut dissimilis, aut saepe verba alia, ratio eadem, et nonnunquam ratio alia, verba eadem. Quae ratio in amor amori, eadem in dolor dolori, neque eadem in dolor dolorem, et cum eadem ratio quae est in amor et amoris sit in amores et amorum, tamen ea, quod non in ea qua oportet confertur materia, per se solum efficere non potest analogias propter disparilitatem vocis figurarum, quod verbum copulatum singulare cum multitudine: ita cum est pro portione, ut eandem habeat rationem, tum denique ea ratio conficit id quod postulat analogia; de qua deinceps dicam.

Of which each several ratio, when compared with another, is either similar or dissimilar, and often the words are different, the ratio the same, and sometimes the ratio different, the words the same. The ratio which is in amor amori is the same in dolor dolori, and not the same in dolor dolorem; and although the same ratio which is in amor and amoris is in amores and amorum, nevertheless that, because it is not brought together in the material in which it ought, by itself alone cannot effect analogies on account of the disparity of the figures of voice, because the coupled word is a singular with a multitude: thus when it is in due proportion, so that it may have the same ratio, then at length that ratio brings about that which analogy demands; about which I shall speak next.

III.

Sequitur tertius locus, quae sit ratio pro portione; ea Graece vocatur ana logon; ab analogo dicta analogia. Ex eodem genere quae res inter se aliqua parte dissimiles rationem habent aliquam, si ad eas duas alterae duae res allatae sunt, quae rationem habeant eandem, quod ea verba bina habent eundem logon, dicitur utrumque separatim analogon, simul collata quattuor analogia.

The third topic follows, what a ratio according to proportion is; this in Greek is called ana logon; from analogon the term analogy is named. From the same kind, things which are in some part dissimilar among themselves have some ratio; if to those two other two things are brought which have the same ratio—because those two words have the same logon—each is called separately an analogon; when the four are compared together, an analogy.

Nam ut in geminis, cum similem dicimus esse Menaechmum Menaechmo, de uno dicimus; cum similitudinem esse in his, de utroque: sic cum dicimus eandem rationem habere assem ad semissem quam habet in argento libella ad simbellam, quid sit analogon ostendimus; cum utrubique dicimus et in aere et in argento esse eandem rationem, tum dicimus de analogia.

For as with twins, when we say that Menaechmus is similar to Menaechmus, we speak about one; when we say that there is similarity in these, about both: so, when we say that the as has the same ratio to the semis as in silver the libella has to the simbella, we show what an analogon is; when we say in both places, both in bronze and in silver, that there is the same ratio, then we speak about analogia.

Ut sodalis et sodalitas, civis et civitas non est idem, sed utrumque ab eodem ac coniunctum, sic analogon et analogia idem non est, sed item est congeneratum. Quare si homines sustuleris, sodalis sustuleris; si sodalis, sodalitatem: sic item si sustuleris logon, sustuleris analogon; si id, analogian.

Just as sodalis and sodality, citizen and civitas are not the same, but each is from the same source and connected, so analogon and analogia are not the same, but likewise are of the same kind. Wherefore, if you remove men, you have removed the sodalis; if the sodalis, the sodality: so likewise, if you remove the logon, you have removed the analogon; if that, the analogia.

Quae cum inter se tanta sint cognatione, debebis suptilius audire quam dici expectare, id est cum dixero quid de utroque et erit commune, ne expectes, dum ego in scribendo transferam in reliquum, sed ut potius tu persequare animo.

Since these things are with so great a cognation among themselves, you will have to listen more subtly rather than expect it to be said; that is, when I have said something about either one and it will be common, do not wait until I, in writing, transfer it to the remaining one, but rather do you yourself pursue it in mind.

Haec fiunt in dissimilibus rebus, ut in numeris si contuleris cum uno duo, sic cum decem viginti: nam quam rationem duo ad unum habent, eandem habent viginti ad decem; in nummis in similibus sic est ad unum victoriatum denarius, sicut ad alterum victoriatum alter denarius; sic item in aliis rebus omnibus pro portione dicuntur ea, in quo est sic quadruplex natura, ut in progenie quomodo est filius ad patrem, est filia ad matrem, et ut est in temporibus meridies ad diem, sic media nox ad noctem.

These things occur in dissimilar matters, as in numbers: if you compare with one two, so with ten twenty; for the same ratio which two have to one, twenty have to ten; in coins, in like cases, thus a denarius is to one victoriatus, just as to another victoriatus another denarius; so likewise in all other things they are said in proportion, in which there is thus a quadruple nature, as in progeny: as a son is to a father, so a daughter is to a mother; and as in times midday is to the day, so midnight to the night.

Hoc poetae genere in similitudinibus utuntur multum, hoc acutissime geometrae, hoc in oratione diligentius quam alii ab Aristarcho grammatici, ut cum dicuntur pro portione similia esse amorem amori, dolorem dolori, cum ita dissimile esse videant amorem et amori, quod est alio casu, item dolorem dolori, sed dicunt, quod ab similibus.

This kind the poets use much in similitudes, this the geometers most acutely, this in oration more diligently than others the grammaticians from Aristarchus, as when things are said, according to proportion, to be similar—love to love, pain to pain—although they see "love" and "to love" to be so dissimilar, because it is in another case, likewise "pain" to "pain"; but they say it on the ground of similars.

Nonnunquam rationes habet implicatas duas, ut sit una derecta, altera transversa. Quod dico, apertius sic fiet. Esto sic expositos esse numeros, ut in primo versu sit unum duo quattuor, in secundo decem viginti quadraginta, in tertio centum ducenti quadringenti.

Sometimes it has two implicated ratios, so that one is direct, the other transverse. What I say will be made clearer thus. Let the numbers be set out thus, that in the first line there be one two four, in the second ten twenty forty, in the third one hundred two hundred four hundred.

In hac formula numerorum duo inerunt quos dixi logoe, qui diuersas faciant analogias: unus duplex qui est in obliquis versibus, quod est ut unus ad duo, sic duo ad quattuor; alter decemplex in directis ordinibus, quod est ut unum ad decem, sic decem ad centum.

In this formula of numbers there will be two logoi which I mentioned, which make diverse analogies: one double, which is in the oblique verses, namely as one to two, so two to four; the other tenfold, in the direct orders, namely as one to ten, so ten to one hundred.

Similiter in verborum declinationibus est bivium, quod et ab recto casu declinantur in obliquos et ab recto casu in rectum, ita ut formulam similiter efficiant, quod sit primo versu hic albus, huic albo, huius albi, secundo haec alba, huic albae, huius albae, tertio hoc album, huic albo, huius albi. Itaque fiunt per obliquas declinationes ex his analogiae hoc genus Albius Atrius, Albio Atrio, quae scilicet erit particula ex illa binaria, per directas declinationes Albius Atrius, Albia Atria, quae scilicet denaria, formula analogiarum, de qua supra dixi.

Similarly, in the declensions of words there is a bivium, since they are declined both from the straight case into the oblique cases and from the straight case into the straight, so that they likewise bring about the formula, namely that in the first line it is hic albus, huic albo, huius albi, in the second haec alba, huic albae, huius albae, in the third hoc album, huic albo, huius albi. Thus, through oblique declensions from these there are formed analogies of this kind Albius Atrius, Albio Atrio, which, of course, will be a particle from that binary one; through direct declensions Albius Atrius, Albia Atria, which, of course, [is] the denary formula of analogies, about which I said above.

Analogia quae dicitur, eius genera sunt duo: unum deiunctum sic est: ut unum ad duo sic decem ad viginti; alterum coniunctum sic: ut est unum ad duo, sic duo ad quattuor. In hoc quod duo bis dicuntur et tum cum conferimus ad unum et tunc cum ad quattuor, hoc quoque natura dicitur quadruplex,; sic e septem chordis citharae tamen duo dicuntur habere tetrachorda, quod quemadmodum crepat prima ad quartam chordam, sic quarta ad septumam respondet, media est alterius prima, alterius extrema. Medici in aegroto septumos dies qui observant, quarto die ideo diligentius signa morbi advertunt, quod quam rationem habuit primus dies ad quartum eandem praesagit habiturum qui est futurus ab eo quartus, qui est septumus a primo.

Analogy, which is so called, has two genera: one disjoined is thus: as one to two, so ten to twenty; the other conjoined thus: as one is to two, so two to four. In this, that “two” is said twice, both when we compare to one and when to four, this too by nature is called quadruple,; thus from the seven chords of the cithara nevertheless they are said to have two tetrachords, because just as the first answers to the fourth chord, so the fourth answers to the seventh: the middle is the first of the one, the extreme of the other. Physicians, who observe the seventh days in a sick person, on the fourth day for that reason more diligently notice the signs of the disease, because the same proportion which the first day had to the fourth it presages that the one which is going to be the fourth from it, which is the seventh from the first, will have.

Quadruplices deiunctae in casibus sunt vocabulorum, ut rex regi, lex legi, coniunctae sunt triplices in verborum tribus temporibus, ut legebam lego legam, quod quam rationem habet legebam ad lego hanc habet lego ad legam. In hoc fere omnes homines peccant, quod perperam in tribus temporibus haec verba dicunt, cum proportione volunt pronuntiare.

The disjoined are fourfold in the cases of words, as rex to regi, lex to legi; the conjoined are threefold in the three tenses of verbs, as legebam, lego, legam, because the same ratio that legebam has to lego, this lego has to legam. In this nearly all men err, in that they improperly say these words in the three tenses when they wish to pronounce with proportion.

Nam cum sint verba alia infecta, ut lego et legis, alia perfecta, ut legi et legisti, et debeant sui cuiusque generis in coniungendo copulari, et cum recte sit ideo lego ad legebam, non recte est lego ad legi, quod legi significat quod perfectum: ut haec tutudi pupugi, tundo pungo, tundam pungam, item necatus sum verberatus sum, necor verberor, necabor verberabor, iniuria reprehendant, quod et infecti inter se similia sunt et perfecti inter se, ut tundebam tundo tundam et tutuderam tutudi tutudero; sic amabar amor amabor, et amatus eram amatus sum amatus ero. Itaque inique reprehendunt qui contra analogias dicunt, cur dispariliter in tribus temporibus dicantur quaedam verba, natura cum quadruplex sit analogia.

For since some verbs are of the unfinished series, as lego and legis, others of the perfected, as legi and legisti, and since each ought, in conjoining, to be coupled with its own kind, and since therefore it is correct to pair lego with legebam, it is not correct to pair lego with legi, because legi signifies what is perfect: as these, tutudi pupugi, tundo pungo, tundam pungam; likewise necatus sum verberatus sum, necor verberor, necabor verberabor, they find fault unjustly, because both the unfinished are like among themselves and the perfect among themselves, as tundebam tundo tundam and tutuderam tutudi tutudero; so amabar amor amabor, and amatus eram amatus sum amatus ero. And so they censure unfairly who, contrary to the analogies, say why certain verbs are said dissimilarly in the three times, since analogy by nature is fourfold.

Id nonnunquam, ut dixi, pauciores videtur habere partes, sic etiam alias pluris, ut cum est: quemadmodum ad tria unum et duo, sic ad sex duo et quattuor, quae tamen quadripertito comprehenditur forma, quod bina ad singula conferuntur; quod in oratione quoque nonnunquam reperietur sic: ut Diomedes confertur Diomedi et Diomedis, sic dicitur ab Hercules Herculi et Herculis.

That sometimes, as I have said, seems to have fewer parts, likewise at other times more, as when it is thus: just as to three belong one and two, so to six two and four, which nevertheless is comprehended in a quadripartite form, because pairs are matched to singles; which also in discourse will sometimes be found thus: as Diomedes is matched to Diomedi and Diomedis, so from Hercules are said Herculi and Herculis.

Et ut haec ab uno capite ac recto casu in duo obliquos discedunt casus, sic contra multa ab duobus capitibus recti casuis confluunt in obliquom unum. Nam ut ab his rectis hi Baebiei, hae Baebiae fit his Baebieis, sic est ab his hi Caelii, hae Caeliae his Caeliis. A duobus similibus dissimiliter declinantur, ut fit in his nemus holus, nemora holera.

And just as these from one head and the straight (nominative) case depart into two oblique cases, so conversely many from two heads of straight cases flow together into a single oblique. For as from these nominatives hi Baebiei, hae Baebiae there is made his Baebieis, so from these hi Caelii, hae Caeliae there is his Caeliis. From two similar ones they are declined dissimilarly, as happens in these: nemus, holus; nemora, holera.

Alia ab dissimilibus similiter declinantur, ut in articulis ab hic iste, hunc istunc.

Other [forms] from dissimilars are declined similarly, as in the articles: from hic, iste; from hunc, istunc.

Analogia fundamenta habet aut a voluntate hominum aut a natura verborum aut a re utraque. Voluntatem dico impositionem vocabulorum, naturam declinationem uocabulorum, quo decurritur sine doctrina. Qui impositionem sequetur, dicet, si simile in recto casu dolus et malus, fore in obliquo dolo et malo; qui naturam sequetur, si sit simile in obliquis Marco Quinto, fore ut sit Marcus Quintus; qui utrumque sequetur, dicet si sit simile, transitus ut est in servus serve, fore ut sit item cervus cerve.

Analogy has its foundations either from the will of men or from the nature of words or from both. By “will” I mean the imposition of vocables; by “nature” the declension of vocables, along which one proceeds without doctrine. He who follows the imposition will say, if “dolus” and “malus” are similar in the nominative case, that in the oblique it will be “dolo” and “malo”; he who follows nature, if it be similar in the obliques “Marco Quinto,” that it will be “Marcus Quintus”; he who follows both will say, if it be similar, that the transitions, as they are in “servus serve,” will likewise be “cervus cerve.”

Commune omnium est, ut quattuor figurae vocis habeant proportione declinatus.

It is common to all, that the four figures of voice have declensions in proportion.

Primum genus est ortum ab similitudine in rectis casibus, secundum ab similitudine quae est in obliquis, tertium ab similitudine quae est in transitibus de casu in casum. Primo genere ab imposito ad naturam proficiscimur, in secundo contra, in tertio ab utroque. Quocirca etiam hoc tertium potest bifariam divisum tertium et quartum dici, quod in eo vel prosus et rusus potest dici.

The first kind takes its origin from likeness in the direct cases, the second from the likeness which is in the obliques, the third from the likeness which is in the transitions from case to case. In the first kind we proceed from imposition to nature; in the second, the other way around; in the third, from both. Wherefore this third too, divided in twofold fashion, can be called the third and the fourth, because in it one can speak either forward (prosus) or backward (rusus).

Qui initia faciet analogiae impositiones, ab his obliquas figuras declinare debebit; qui naturam, contra; qui ab utraque, reliquas declinationes ab eiusmodi transitibus. Impositio est in nostro dominatu, nos in naturae: quemadmodum enim quisque volt, imponit nomen, at declinat, quemadmodum volt natura.

He who will set the beginnings of analogy in impositions will have to decline the oblique figures from these; he who (sets them) in nature, the contrary; he who from both, the remaining declensions from transitions of such a kind. Imposition is in our dominion, we in that of nature: for just as each person wills, he imposes a name, but he declines as nature wills.

Sed quoniam duobus modis imponitur vocabulum aut re singulari aut multitudine, singulari, ut cicer, multitudinis, ut scalae, nec dubium est, quin ordo declinatuum, in quo res singulares declinabuntur solae, ab singulari aliquo casu proficiscatur, ut cicer ciceri ciceris, item contra in eo ordine, qui multitudinis erit solum, quin a multitudinis aliquo casu ordiri conueniat, ut scalae scalis scalas: aliud videndum est, cum duplex natura copulata ac declinatuum bini fiant ordines, ut est Mars Martes, unde tum ratio analogiae debeat ordiri, utrum ab singulari re in multitudinem an contra.

But since a vocable is imposed in two ways, either upon a singular thing or upon a multitude—on a singular, as cicer, on a multitude, as scalae—there is no doubt that the order of declensions, in which only singular things will be declined, sets out from some singular case, as cicer ciceri ciceris; likewise, conversely, in that order which will belong to the multitude alone, that it is fitting to begin from some case of the multitude, as scalae scalis scalas: something else must be considered, when a double nature is coupled and two orders of declensions are made, as is Mars Martes, whence then the ratio of analogy ought to take its beginning, whether from the singular thing into the multitude or the contrary.

Neque enim si natura ab uno ad duo pervenit, idcirco non potest apertius esse in docendo posterius, ut inde incipias, ut quid sit prius ostendas. Itaque et hi qui de omni natura disputant atque ideo vocantur physici, tamen ex his ab universa natura profecti retro quae essent principia mundi ostendunt. Oratio cum ex litteris constet, tamen ex ea grammatici de litteris ostenderunt.

Nor indeed, if nature comes from one to two, does it therefore follow that the later cannot be more open in teaching, so that you begin from there in order to show what is earlier. And so even those who dispute about all nature and are therefore called physicists (natural philosophers), having set out from universal nature, show retrospectively what the principles of the world were. Discourse, although it consists of letters, nevertheless from it the grammarians have demonstrated concerning letters.

Quare in demonstrando, quoniam potius proficisci oportet ab eo quod apertius est quam ab eo quod prius est et potius quam a corrupto principio ab incorrupto, ab natura rerum quam ab lubidine hominum, et Haec tria quae sequenda magis sunt minus sunt in singularibus quam in multitudinis, a multitudine commodius potest ordiri, quod in his principiis minus rationis verbis fingendis. Verborum formas facilius ex multitudinis singularis videri posse quam ex singularibus multitudinis haec ostendunt: trabes trabs, duces dux.

Wherefore in demonstrating, since it is more fitting to set out from that which is more open than from that which is prior, and rather from an incorrupt principle than from a corrupt one, from the nature of things than from the desire of men, and since these three which are to be followed more are found less in singulars than in those of the multitude, one can more conveniently begin from the multitude, because in these principles there is less of reason in the fashioning of words. That the forms of words can more easily be seen from the multitude to the singular than from singulars to the multitude, these show: trabes trabs, duces dux.

Videmus enim ex his verbis trabes duces de extrema syllaba E litteram exclusam et ideo in singulari factum esse trabs dux. Contra ex singularibus non tam videmus quemadmodum facta sint ex B et S trabs et ex C et S dux.

For we see from these words trabes duces that the letter E has been excluded from the final syllable and therefore in the singular it has been made trabs dux. Conversely, from the singulars we do not so much see how trabs has been made from B and S, and dux from C and S.

Si multitudinis rectus casus forte figura corrupta erit, id quod accidit raro, prius id corrigemus quam inde ordiemur; ab obliquis adsumere oportet figuras eas quae non erunt ambiguae, sive singulares sive multitudinis, ex quibus id, cuius modi debent esse, perspici possit.

If the straight case of the multitude should by chance have a corrupt figure—which happens rarely—we will first correct that before we begin from it; from the oblique cases one ought to assume those figures which will not be ambiguous, whether singular or of the multitude, from which it can be perceived of what sort they ought to be.

Nam nonnunquam alterum ex altero videtur, ut Chrysippus scribit, quemadmodum pater ex filio et filius ex patre, neque minus in fornicibus propter sinistram dextra stat quam propter dextram sinistra. Quapropter et ex rectis casibus obliqui et ex obliquis recti et ex singularibus multitudinis et ex multitudinis singulares nonnunquam recuperari possunt.

For sometimes the one is seen from the other, as Chrysippus writes, just as the father from the son and the son from the father, nor in arches does the right-hand stand because of the left any less than the left because of the right. Wherefore both from the direct cases the oblique and from the oblique the direct and from the singulars the plural and from the plural the singulars can sometimes be recovered.

Principium id potissimum sequi debemus, ut in eo fundamentum sit, natura, quod in declinationibus ibi facilior ratio. Facile est enim animadvertere, peccatum magis cadere posse in impositiones eas quae fiunt plerumque in rectis casibus singularibus, quod homines imperiti et dispersi vocabula rebus imponunt, quocumque eos libido invitavit: natura incorrupta plerumque est suapte sponte, nisi qui eam usu inscio deprauabit.

We ought most of all to follow this principle, that the foundation be nature, because in declensions the rationale is easier there. For it is easy to observe that error can fall more upon those impositions which are made for the most part in the right cases of the singular, because unskilled and scattered men impose vocables upon things wherever their fancy has invited them: nature, uncorrupted, is for the most part of its own accord, unless someone through unknowing use will deprave it.

Quare si quis principium analogiae potius posuerit in naturalibus casibus quam in impositiciis, non multa inconcinna in consuetudine occurrent et a natura libido humana corrigetur, non a libidine natura, quod qui impositionem sequi voluerint facient contra.

Wherefore, if anyone should set the principle of analogy rather in natural cases than in imposititious ones, not many incongruities will occur in usage, and human caprice will be corrected by nature, not nature by caprice—whereas those who will have wished to follow imposition will do the contrary.

Sin ab singulari quis potius proficisci volet, initium facere oportebit ab sexto casu, qui est proprius Latinus: nam eius casuis litterarum discriminibus facilius reliquorum varietatem discernere poterit, quod ei habent exitus aut in A, ut hac terra, aut in E, ut hac lance, aut in I, ut hac clavi, aut in O, ut hoc caelo, aut in U, ut hoc versu. Igitur ad demonstrandas declinationes biceps via haec.

But if anyone will rather set out from the singular, it will be proper to make the beginning from the sixth case, which is peculiarly Latin: for by the distinctions of the letters in its case-forms he will be able more easily to discern the variety of the rest, because they have endings either in A, as in hac terra, or in E, as in hac lance, or in I, as in hac clavi, or in O, as in hoc caelo, or in U, as in hoc versu. Therefore, for demonstrating the declensions, this way is twofold.

Sed quoniam ubi analogia, tria, unum quod in rebus, alterum quod in vocibus, tertium quod in utroque, duo priora simplicia, tertium duplex, animadvertendum haec quam inter se habeant rationem.

But since where there is analogy, there are three, one which is in things, another which is in voices, a third which is in both, the two former simple, the third double, it must be observed what relation these have among themselves.

Primum ea quae sunt discrimina in rebus, partim sunt quae ad orationem non attineant, partim quae pertineant. Non pertinent ut ea quae observant in aedificiis et signis faciendis ceterisque rebus artifices, e quis vocantur aliae harmonicae, sic item aliae nominibus aliis: sed nulla harum fit in loquendo pars.

First, those discriminations which are in things are partly such as do not pertain to oration, partly such as do pertain. They do not pertain, as those which artificers observe in edifices and in making signs (statues) and in other things; from which some are called harmonics, likewise others by other names: but none of these becomes a part in speaking.

Ad orationem quae pertinent, res eae sunt quae verbis dicuntur pro portione neque a similitudine quoque vocum declinatus habent, ut Iupiter Marspiter, Iovi Marti. Haec enim genere nominum et numero et casibus similia sunt inter se, quod utraque et nomina sunt et virilia sunt et singularia et casu nominandi et dandi.

Those which pertain to speech are those things which are said by words in proportion and do not have inflections deviating from the likeness of the sounds as well, as Jupiter Marspiter, to Jove to Mars. For these are similar among themselves in the gender of the nouns and in number and in cases, because both are nouns and are masculine and singular and in the case of naming and of giving.

Alterum genus vocale est, in quo voces modo sunt pro portione similes, non res, ut biga bigae, nuptia nuptiae: neque enim in his res singularis subest una, cum dicitur biga quadriga, neque ab his vocibus quae declinata sunt, multitudinis significant quicquam, id quod omnia multitudinis quae declinantur ab uno, ut a merula merulae: sunt enim eius modi, ut singulari subiungatur, sic merulae duae, catulae tres, faculae quattuor.

The other kind is vocal, in which the words are only proportionally similar, not the things, as biga bigae, nuptia nuptiae: for in these no single thing underlies when one says biga, quadriga, nor do they, from these words when declined, signify anything of a multitude, which all pluralities that are declined from one do, as from merula to merulae: for they are of such a sort that a numeral is subjoined, thus, two blackbirds, three whelps, four little torches.

Quare cum idem non possit subiungi, quod non dicimus biga una, quadrigae duae, nuptiae tres, sed pro eo unae bigae, binae quadrigae, trinae nuptiae, apparet non esse a biga et quadriga bigae et quadrigae, sed ut est huius ordinis una duae tres principium una, sic in hoc ordine altero unae binae trinae principium est unae.

Wherefore, since the same cannot be subjoined—for we do not say biga una, quadrigae duae, nuptiae tres, but instead unae bigae, binae quadrigae, trinae nuptiae—it is apparent that bigae and quadrigae are not from biga and quadriga; but just as in that order “one two three” the beginning is una, so in this other order “unae binae trinae” the beginning is unae.

Tertium genus est illud duplex quod dixi, in quo et res et voces similiter pro portione dicuntur ut bonus malus, boni mali, de quorum analogia et Aristophanes et alii scripserunt. Etenim haec denique perfecta ut in oratione, illae duae simplices inchoatae analogiae, de quibus tamen separatim dicam, quod his quoque utimur in loquendo.

The third genus is that double one which I said, in which both things and words are said similarly in proportion, as “good” and “bad,” “good ones” and “bad ones,” about the analogy of which both Aristophanes and others have written. For indeed these are, in fine, perfect, as in discourse; those two, however, are simple, inchoate analogies, about which I shall speak separately, because we too make use of these in speaking.

Sed prius de perfecta, in qua et res et voces quadam similitudine continentur, cuius genera sunt tria: unum vernaculum ac domi natum, alterum adventicium, tertium nothum ex peregrino hic natum. Vernaculum est ut sutor et pistor, sutori pistori; adventicium est ut Hectores Nestores, Hectoras Nestoras; tertium illum nothum ut Achilles et Peles.

But first about the perfect kind, in which both things and words are contained by a certain similitude, of which the genera are three: one vernacular and born at home, another adventitious, a third nothus, born here from a foreigner. The vernacular is as sutor and pistor, sutori pistori; the adventitious is as Hectores Nestores, Hectoras Nestoras; the third, that nothus, as Achilles and Peles.

De his primo genere multi utuntur non modo poetae, sed etiam plerique omnes qui soluta oratione loquuntur. Haec primo dicebant ut quaestorem praetorem, sic Hectorem Nestorem: itaque Ennius ait:

Of these, the first kind many make use, not only poets, but also very many of all who speak in prose. These at first they used to say as “quaestorem praetorem,” thus “Hectorem Nestorem”: and so Ennius says:

Hectoris natum de muro iactarier.

Hector’s son to be hurled from the wall.

Accius haec in tragoediis largius a prisca consuetudine movere coepit et ad formas Graecas verborum magis revocare, a quo Valerius ait:

Accius in tragedies began to set these things in motion more liberally than the ancient usage and to recall words more to Greek forms, by whom Valerius says:

Accius Hectorem nollet facere, Hectora mallet.

Accius would not wish to make Hectorem, he would prefer Hectora.

Quod adventicia pleraque habemus Graeca, secutum ut de nothis Graecanicos quoque nominatus plurimos haberemus. Itaque ut hic alia Graeca, alia Graecanica, sic analogiae.

Because most of our adventitious things we have are Greek, it followed that among the bastard [forms] we should have very many Graecanic appellations as well. And so, as here some are Greek, others Graecanic, so too the analogies.

E quis quae hic nothae fiunt declinationes, de his aliae sunt priscae, ut Bacchides et Chrysides, aliae iuniores, ut Chrysides et Bacchides, aliae recentes, Chrysidas et Bacchidas; cum his omnibus tribus utantur nostri, maxime qui sequontur media in loquendo offendunt minimum, quod prima parum similia videntur esse Graecis, unde sint tralata, tertia parum similia nostris.

And of these, the nothoi declensions that are made here, of these some are pristine, as Bacchides and Chrysides, others junior, as Chrysides and Bacchides, others recent, Chrysidas and Bacchidas; since our people use all three of these, those who follow the middle in speaking offend the least, because the first seem too little similar to the Greeks, whence they have been translated, the third too little similar to our own.

IV.

Omnis analogiae fundamentum similitudo quaedam, ea, ut dixi, quae solet esse in rebus et in vocibus et in utroque; in quam harum partem quodque sit inferendum e cuius modi, videndum. Nam, ut dixi, neque rerum neque vocis similitudo ad has duplicis quas in loquendo quaerimus analogias verborum exprimendas separatim satis est, quod utraque parte opus est simili. Quas ad loquendum ut perducas accedere debet usus: alia enim ratio qua facias vestimentum, alia quemadmodum utare vestimento.

The foundation of every analogy is a certain similarity, that, as I said, which is wont to be in things and in words and in both; it must be considered into which of these parts each is to be brought, and of what sort. For, as I said, neither the similarity of things nor of word is separately sufficient for expressing those double analogies of words which we seek in speaking, since there is need of a like on both sides. In order that you may bring these through to speaking, usage ought to be added: for one rationale is that by which you make a vestment, another how you use the vestment.

Usuis species videntur esse tres: una consuetudinis veteris, altera consuetudinis huius, tertia neutrae. Vetera, ut cascus casci, surus suri; huius consuetudinis, ut albus caldus, albo caldo; neutrae, ut scala scalam, phalera phaleram. Ad quas accedere potest quarta mixta, ut amicitia inimicitia, amicitiam inimicitiam.

The species of usage seem to be three: one of ancient custom, another of this present custom, a third of neither. The ancient ones, as cascus casci, surus suri; of this custom, as albus caldus, albo caldo; of neither, as scala scalam, phalera phaleram. To which a fourth, mixed, can be added, as amicitia inimicitia, amicitiam inimicitiam.

Prima est qua usi antiqui et nos reliquimus, secunda qua nunc utimur, tertia qua utuntur poetae.

The first is that which the ancients used and we have left behind; the second is that which we now use; the third is that which the poets use.

Analogia non item ea definienda quae derigitur ad naturam verborum atque illa quae ad usum loquendi. Nam prior definienda sic: analogia est verborum similium declinatio similis, posterior sic: analogia est verborum similium declinatio similis non repugnante consuetudine communi. At quom harum duarum ad extremum additum erit hoc "ex quadam parte," poetica analogia erit definita.

Analogy is not to be defined in the same way in that which is directed to the nature of words and in that which is directed to the usage of speech. For the former is to be defined thus: analogy is a similar declension of similar words; the latter thus: analogy is a similar declension of similar words, with common usage not opposing. But when to the end of these two there shall have been added this—“in some respect”—the poetic analogy will have been defined.

Harum primam sequi debet populus, secundam omnes singuli e populo, tertiam poetae.

Of these, the first the populace ought to follow, the second all the individual members of the populace, the third the poets.

Haec diligentius quam apertius dicta esse arbitror, sed non obscurius quam de re simili definitiones grammaticorum sunt, ut Aristeae, Aristodemi, Aristocli, item aliorum, quorum obscuritates eo minus reprehendendae, quod pleraeque definitiones re incognita propter summam brevitatem non facile perspiciuntur, nisi articulatim sunt explicatae.

I judge that these things have been said more diligently than openly, but not more obscurely than are the definitions of the grammarians on a similar matter, such as those of Aristaeus, Aristodemus, Aristocles, and likewise of others; whose obscurities are so much the less to be reprehended, because most definitions, with the matter unknown, on account of extreme brevity are not easily perceived, unless they have been explained article by article.

Quare magis apparebit, si erit aperte de singulis partibus, quid dicatur verbum, quid similitudo verbi, quid declinatio, quid similitudo declinationis non repugnante consuetudine communi, quid ex quadam parte.

Wherefore it will appear more apparent, if there is an open account of the individual parts: what is called a word, what the similitude of a word, what declension, what the similitude of declension, with the common usage not opposing, and what in a certain respect.

Verbum dico orationis vocalis partem, quae sit indivisa et minima. Si declinationem naturalem habeat, simile verbum verbo tum quom, et re quam significat et voce qua significat et in figura e transitu declinationis parile. Declinatio est, cum ex verbo in verbum aut ex verbi discrimine, ut transeat mens, vocis commutatio fit aliqua.

By “word” I mean a part of vocal speech, which is individed and minimal. If it have a natural declension, a word is similar to a word then when both in the thing which it signifies and in the voice by which it signifies and in the figure, equal in the transition of declension. Declension is, when from word into word, or from a word’s discrimination, so that the mind may pass over, some commutation of the voice is made.

Similitudo declinationis, cum item ex aliqua figura in figuram transit, ut id transit, cum quo confertur.

A similarity of declension is when likewise it passes from some figure into a figure, as that passes with which it is compared.

Adiectum est "non repugnante consuetudine communi", quod quaedam verba contra usum veterem inclinata patietur, ut passa Hortensium dicere pro hae cervices cervix, quaedam non, ut si dicas pro fauces faux. Ubi additur "ex quadam parte", significat non esse in consuetudine in his verbis omnis partis, ut declinatum ab amo vivo amor sed non vivor.

It is added, “with the common custom not resisting,” because certain words will allow being inclined against the old usage—as Hortensius allowed saying “cervix” in place of “hae cervices”—and certain ones will not, as if you were to say “faux” for “fauces.” Where “in a certain part” is added, it signifies that not the whole of the parts is in use in these words, as there is a declined form from “amo,” “amor,” but not from “vivo,” “vivor.”

Quid videretur analogia in oratione et quas haberet species et quae de his sequendae viderentur, ut brevi potui informavi; nunc, in quibus non debeat esse ac proinde ac debeat soleat quaeri, dicam. Ea fere sunt quattuor genera: primum in id genus verbis quae non declinantur analogia non debet quaeri, ut in his nequam mox vix.

What analogy in speech would seem to be, and what species it would have, and which of these would seem to be to be followed, I have, as briefly as I could, set forth; now I will say in what cases it ought not to be, and accordingly in what cases it ought to be sought. These are about four kinds: first, in that kind of words which are not declined, analogy ought not to be sought, as in these: nequam, mox, vix.

De his magis in alio quam in alio erratur verbo. Dant enim non habere casus mox et vix, nequam habere, quod dicamus hic nequam et huius nequam et huic nequam. Cum enim dicimus hic nequam et huius nequam, tum hominis eius, quem volumus ostendere esse nequam, dicimus casus, et ei proponimus tum hic nomen, cuius putamus nequitiam.

Concerning these, one errs more in one word than in another. For they grant that mox and vix do not have cases, that nequam does have them, since we say hic nequam and huius nequam and huic nequam. For when we say hic nequam and huius nequam, then of that man whom we wish to show to be nequam we are speaking the cases, and to him we then set forth this name, whose nequity we suppose.

Quod vocabulum factum ut ex non et volo nolo sic ex ne et quicquam item media extrita syllaba coactum est nequam. Itaque ut eum quem putamus esse non hili dicimus nihili, sic in quo putamus esse ne quicquam dicimus nequam.

That vocable was formed just as from non and volo comes nolo, so from ne and quicquam likewise, with the middle syllable rubbed away, it was compressed into nequam. And thus, just as we say nihili of him whom we think to be non hili, so of one in whom we suppose there to be ne quicquam we say nequam.

Secundo, si unum solum habent casum in voce, quod non declinentur, ut litterae omnes. Tertio, si singularis est vocabuli series neque habet cum qua comparari possit, ut esse putant caput capiti capitis capite. Quartum, si ea vocabula quattuor quae conferuntur inter se rationem non habent quam oportet, ut socer socrus, soceros socrus.

Secondly, if they have only a single case in form, because they are not declined, as with all the letters. Thirdly, if the series of a word is singular and has nothing with which it can be compared, as they think to be the case: caput capiti capitis capite. Fourthly, if those four words which are brought together in comparison with one another do not have the relation which they ought, as socer socrus, soceros socrus.

VI.

Contra in quibus debeat quaeri analogia, fere totidem gradus debent esse coniuncti: primum ut sint res, secundum ut earum sit usus, tertium uti hae res vocabula habeant, quartum ut habeant declinatus naturalis. De primo gradu, quod natura subest et multitudinis et singularis, dicimus hi asses hosce asses, hic as hunc assem; contra quod in numeris finitis multitudinis natura singularis non est, dicitur hi duo et hi tres, his duobus et his tribus.

On the contrary, in those matters in which analogy ought to be sought, there ought to be nearly the same number of connected steps: first, that they be things; second, that there be a use of them; third, that these things have vocables; fourth, that they have natural declensions. About the first step, because nature underlies both plurality and singularity, we say hi asses hosce asses, hic as hunc assem; conversely, because in finite numbers the nature of plurality has no singular, it is said hi duo and hi tres, his duobus and his tribus.

Secundo gradu si est natura neque est usus, id genus ut sit discriminandum, ut fit in faba et id genus, quae item et ex parte et universa nominamus non enim opus fuit ut in servis ***

In the second degree, if there is nature and there is not use, that kind should be discriminated, as happens in the bean and the like, which likewise we name both by part and as a whole; for there was no need as with slaves ***