No CrossRef data available.
Published online by Cambridge University Press: 22 April 2026
The significance of our reconstructed forms, precisely what is intended by them, is an old question which some of us discussed briefly at the St. Louis Congress of Arts and Sciences of 1904 (cf. the publication, vol. III, pp. 35, 57), and which has been elaborated in the long article of E. Hermann, KZ 41. 1 ff. There is not only a difference in the views expressed by various scholars, but often in the case of the same one an apparent discrepancy between the interpretation of reconstruction that is explicitly professed and that which must be inferred from his practice. For example Meillet, Introd. 24 ff., insists on the unreality and the purely formulaic character of the reconstructions, saying that these are nothing but convenient formulae for given correspondences (similarly Oertel, Lectures 128, and others). Yet throughout the work he is constantly, like any other scholar in the field, asserting or discussing the sound, form, or type that must be assumed for the parent speech to account for given correspondences, and whether this and that type is inherited from the parent speech or an innovation. Again, not believing in the existence of three guttural series in the parent speech, he does not recognize the ‘plain velars’ of other scholars, whereas from his professed principle one would expect him to have the least compunction in admitting them as convenient formulae for a well-known set of correspondences. The fact is, of course, that to him, and to all, the reconstructions, while mainly useful as formulae, are still something more than mere formulae of correspondences, they imply a certain interpretation of these correspondences, a conviction or a provisional theory regarding their approximate common starting point.
1 Opening of a round table discussion at the meeting of the Linguistic Society of America, Dec. 29, 1925.
2 Cf. Kenyon, American Pronunciation, 74, 77, with whose analysis my own pronunciation seems to agree perfectly. There is no reason to believe that British pronunciation is essentially different in this matter.
3 The medial glides may of course be strengthened and become significant element (Cf. Lat. ruina, vidua, Genua to Ital. rovina, vedova, Genova). It is possible that that is true in those languages in which they are regularly represented in writing. But the difference may be purely graphic.
4 Hermann, Gött. Nachr. 1918, attacks current reconstructions with i and
, but from a different point of view. He objects specifically to the sequence i
, u
, as impossible under our understanding of IE
,
as ‘semivowels.‘ On the contrary that is just where in English (ye, woo) the y, w differ most distinctly from the vowels i, u. Hirt of course would not agree that the y, w in such cases are negligible glides, for to him they are retentions of the y, w from the full grades ey, ew, in which the e was reduced to an obscure vowel and then assimilated (Idg. Vok. 87). But the historical forms (e. g. Grk.
, Lat. trium Goth, prijē point to IE i and u respectively, and nothing else. That these ever differed from the i and u before consonants (Grk.
, Lat tribus) or stood in any different relation to the strong grade is a conclusion that cannot be reached by the historical method and which most of us do not accept. Our reconstructions should rest on the historical basis rather than on theories of Proto-IE relations. We should reason back from the known, and not forward from the remote and speculative.
5 To meet a specific objection, namely that
nn should yield in Germanic unn instead of the actual un and similarly in Balto-Slavic inn instead of actual in, Brugmann, Grd.1 2.920, (1892) proposed a revised notation
n etc., and similarly i
, u
to indicate that the glides were not equivalent to the full consonants. He followed this revised form in Grd.2 1. 395ff. (1897), but reverted to
nn etc. in the Kurze vergl. Gram. 122ff. (1902) and in the later volumes of the second edition of the Grundriss.
6 In this case Hirt appeals to the historical representation as always showing vowel +nasal or liquid. But the argument loses its apparent merit when we take into account the fact in the reduction of ne, etc. the result is precisely the same, e. g. the antivocalic form of the negative prefix coming from the full ∗ne (or pronominal ∗
s beside ∗nes). He would doubtless take refuge in a dissylabic ∗∗ene but this would be wholly in the air.
7 No one doubts that Skt. īr, ūr stand in the same relation to
as ū to u or ī to i, and represent a parallel reduction of dissylabic bases. No one doubts that Skt. īr, ūr have certain monosyllabic correspondences in the European languages (the only outstanding dispute being whether Greek ρω, λω or
, with the probability, as I now believe, in favor of the latter). Most will continue to believe that these correspondences reflect a monosyllabic reduced grade in the parent speech, parallel to ū, ī. Hirt sets up for the parent speech only uncontracted dissyllabic forms not only for Skt. īr, ūr, etc., but also for what is ī, ū in all the IE languages. In this last he has, I think, no followers. To doubt that an ī or ū which occurs uniformly in all or most of the IE languages (e. g. the ū of Skt.abhūt, Grk.
, Ch. Sl. by, etc.) reflects IE ī or ū, is to abandon the historical and only safe method of reconstruction. And it is quite unnecessary for the general theory of dissyllabic bases. My interpretation of the situation is virtually the same as that of Per Persson, Beitrāge II 631ff. In the case of Proto-IE ∗∗eyā x, ∗∗ewā x, whence in IE with accent on either syllable ∗ey
, ∗ew
or ∗yā x, ∗wā x respectively, the forms with accent on neither syllable and reduction of each would be, in the notation proposed above, ∗i
, ∗u
(R+R), which were normally contracted to IE ī, ū (RR) as reflected in the historical period. Sometimes however, under special conditions not now determinable, the i
, u
remained uncontracted and are reflected in a few forms like Grk.
= Skt. patnī,
, beside Skt. krī-. Similarly in the case of Proto-IE ∗∗enā x, ∗∗elā x, etc. the forms with reduction of each syllable would be ∗![]()
, ∗l
etc. which again under certain conditions remained uncontracted and are reflected by Grk. ανα, αδα, etc. (
etc.), but were normally contracted to something that is reflected by Skt. ā, ām, īr, ūr, Grk.
, etc., Lat. nā, rā, etc.
Otherwise Meillet, who used
, etc. in his Introduction but in the Traité de grammaire comparée des langues classiques, 156, sets up ∗y
, ∗w
, whence ī, ū, and ∗n
, ∗l
, etc., whence nā, lā, etc. in Greek and Latin, e g. ∗gn
- for Lat. (g) nātus and (g)nārus. But ∗n
is commonly used to represent another probable though rare form of reduction, and should moreover yield European na not nā.