hatela'ah (Exod. 18:8) ('the weariness') Rashi comments, "Lamed and Alef are the basic elements of the word, and the Tav is an 'improvement' and a 'basic element' which sometimes drops off. Words of a similar structure are: teruma, tenufa, tekuma, tenu'ah." Here Rashi seems to distinguish between affixes that have no meaning ('bound empty morphemes') but attach to a root to produce a word with meaning of its own ('independent morpheme'), and affixes ('bound morphemes') that have their own meanings that can be attached to a root to produce a series of words (a conjugation or declension). However an examination of the 14 references in the Tenach where Rashi uses the term 'basic element which sometimes drops off' (provided by the program Chalamish), indicates that he also uses it for letters that later grammarians regarded as letters of the root which are elided (Chaserim, Kefulim). Rashi's examples include affixes of both two- and three-letter roots. This is consistent with the grammatical approach of Menachem, whom Rashi frequently quotes. The fact that Rashi accepts Menachem's approach to Hebrew roots causes difficulty to readers familiar with standard Hebrew grammar, as Menachem's theory was rejected (ignored) by R' David Kimche (Radak) and thereafter by all mainstream grammarians including R' Eliyahu of Vilna (Gr"a). R' Tzvi (Herschl) Filipowski revived interest in Menachem's view, by publishing his Machberet from manuscripts (Edinburgh, 1854). A facsimile print was published in recent years (Vagshal, Jerusalem, nd).

In the two versions of R' A. ibn Ezra's commentary on Exodus that have come down to us, one ('the long') gives Lamed Alef Heh as the root, and the other ('the short') reports some (ie Rashi) give Lamed Alef as the root, with R' A. ibn Ezra commenting that this is unlikely. As to Tav, R' A. ibn Ezra states simply that it is an affix.

Why Does Rashi Add a Midrash?

vayichad (Exod. 18:9) Rashi comments "'And Yitro rejoiced', this is the plain meaning." Here Rashi maintains that according to the plain meaning this word is related to chedva 'happiness'. But then he adds "However the Midrash is 'His flesh became covered with pimples over the destruction of Egypt' (the source of the Midrash is Sanh. 94a)"

R' A. ibn Ezra states "This word is related to chedva and its vowels should have been the same as those in vayift ('and he turned') (Job 31:27), and in vayishb ('and he captured') (Num. 21:1) but the Chet received a Patach because it is a guttural letter, and the Dalet remained with its Dagesh as though the original vowels were there. I do not know of a letter having a Dagesh at the end of a word other than Bet Gimmel Dalet Kaf Peh Tav which occur where there are two Shevas as in vayishb (above), and vayasht (Gen. 9: 21) … whereas following Patach we have only vayichad (here and Job 3:5/6)."

Rashi's first explanation is the same as that of R' A. ibn Ezra. Why does he feel that this is insufficient? Why does Rashi add a Midrash? R' Eliyahu Mizrachi (Ram) answers this question by writing "If so it should have said vayismach as it does everywhere." However, according to the opinion of R' Meir Leibush ben Yechiel Michel (Malbym), no two synonyms have truly identical meaning (this seems also to be the opinion of the Gra), so vayismach and vayichad must be distinguished. Indeed were it not so, why should a language carry the two words? Malbym (ad loc) writes, "It would seem that the difference between chedva and other expressions indicating joy is that chedva denotes inner spiritual joy which overcomes externally generated sadness.

Perhaps in his first comment Rashi accepted the same analysis as R' A. ibn Ezra. However this assumes that there are true synonyms in Hebrew, which to say the least, is a matter for debate (between Ram and Malbym). Furthermore it leaves the Dagesh in the Dalet of vayichad without a parallel as R' A. ibn Ezra himself points out. Therefore Rashi quotes the Midrash, which deals with both these problems. The issue of synonyms is avoided, as the word is now interpreted to mean pointed things ('pimples'), and the verse has a new meaning. The Dagesh in the Dalet of vayichad is seen as indicating the doubling of the Dalet, solving the problem of the Dagesh in one sense, though creating new problems of having a strong Dagesh at the end of a word. This probably means that the Sheva is a sounding Sheva but it has no practical effect, as for purposes of reading we follow the plain meaning.

The verbs hashmein, hachbeid, and hasha (Isaiha 6:10) are all hif'il. In this conjugation the infinitive and the imperative forms are indistinguishable. Hence Rashi and R' A. ibn Ezra base their difference of opinion on context.