Note note 86 — ANALYSIS OF THE PHENOMENA OF THE HUMAN MIND James Mill

86 The abstracts in -th belong to a very early stage of the language. We cannot now form words like health, truth, as we can abstracts in -ness. As in the case of adjectives in -en (wooden), and of preterites and participles like fell, fallen, that particular part of the vital energy of the language that produced them, is dead—ossified, as it were; and we cannot exemplify their formation by any process now going on. To account for many of them, we must suppose them formed from roots different from any now existing as separate words—roots from which the corresponding verbs and adjectives that we are acquainted with have been themselves derived by augmentation or other change. This being the case, it is impossible to say with certainty whether the immediate root of any particular abstract in -th was a verb, a noun, or an adjective; and, indeed, the question need hardly be raised, since a primitive root was of the nature of all three.

The structure of these derivatives is better seen in some of the other Teutonic dialects than in the English or the Anglo-Saxon, in which the affix is reduced to a mere consonant. Thus, for Eng. depth the Gothic has diupi-tha; for heigh-th, hauhi-tha. In Old High German the affix -tha becomes -da, and we have heili-da corresponding to Eng. heal-th; strenki-da, to streng-th; besides a great number of analogous forms, such as evi-da, “eternity” (from the same root as ever; compare Lat. aetas for aevitas). In modern German comparatively few of these derivatives survive; and in those that do; the -da of the Old German has passed into -de, as in ge-baer-de, the way of ‘bearing’ oneself, behaviour; equivalent to Latin habi-tus. The modern German equivalents of bread-th, leng-th, are breit-e, läng-e; but in some of the popular dialects the older forms breite de, läng-de are still retained; and in Dutch warm-te corresponds to warm-th, and grôt-te is great-ness. When we recollect that th or d in the Germanic languages represents in such cases the t of the Greek and Latin (compare Gr. μέλιτ (ος), honey with Goth. milith; Lat. alter with Eng. other), we cannot help seeing how analogous is the formation of the class of words we are now considering to that of Latin past participles (ama-tus, dic-tus, audi-tus). In the case of those abstracts that seem to come more naturally from an adjective root than from a verb, we can conceive the adjective formed on the analogy of the past participle; just as there are in English adjectives having no possible verbal root, yet simulating past participles; as able-bodi-ed, three-corner-ed. The abstract noun would appear to have been originally distinguished from the participle, or participial adjective, by some additional affix, as in lec-t-io. In Greek and Latin this additional affix very often consisted in a reduplication of the formative element t, as if for the purpose of denoting multitude, generality; as in Greek (νεό-τητ-ος), Latin juven-tut-is, sani-tat-is. It is not impossible that Goth. diupi-tha, O.H.G. heili-da are abbreviations of diupi-tha-th, heili-da-d, just as Lat. sani-tat has dwindled down in modern Ital. to sani-tà.

In a great many words essentially belonging to the same class both in meaning and in mode of formation, the -th has, for the sake of euphony or from other causes, given place to t or d. Thus mood corresponds to Goth. mo-th, and means a motion (Lat. motus) or affection (of the mind); blood, to Goth. blo-th; theft, is in Ang. Sax. theof-th. Mur-ther, from a root akin to Lat. mori; burthen, from the root of to bear, are of similar formation, with additional affixes.

All these considerations would seem to put Horne Tooke’s proposed derivation of these abstracts from the third person singular of the present indicative of the verb, completely out of court. The famous case of truth from troweth is especially absurd. For one thing the Ang. Sax. verb treowan does not mean “to think,” but “to trust,” “rely on,” “believe.” This implies a ground for the trust, and that ground lies in the quality expressed by the adjective, true. Truth has the same relation, logically and etymologically, to true, that dearth has to dear, health to hale. Remarking on the identity in form between the Ang. Sax. treow, “trust,” “a treaty,” and treow, “a tree,” Jacob Grimm suggests that they are radically related, and that the idea common to tree and true is firmness, fixedness. Thus the “true” would be the “firm” the “fixed”—what may be relied on. This view is supported by the analogy of the Lat. robur, which means both an oak and strength.—F.