concord (see par. 8, 17).
50/ When two verbs are in consecutive independent clauses and have the same tense and the same subject, the second may be put into the infinitive. (Speakers of European languages seldom catch on to this useful trick.)
51/ What is historically the ku of the infinitive shows up in yet another way in affirmative tenses when the stem of the verb is monosyllabic (-pa'‘give,' -wa 'be, become,' -ja; 'come,' -fa; 'die,' -la 'eat' etc.). The word stress always falls on the next-to-last syllable of a word. But there are certain prefixes which never take word stress. They are egg: , -na- , -li- , -ta-, -me- and -nge- (par. 29 ), and the relative prefixes (par. 42 ). If one of these would otherwise be the next-to-last syllable of the verb, then the meaningless syllable ku is inserted (from a historical point of view, 'retained,' and not 'inserted'). 52/ In the same tenses, if there is an object prefix, the meaningless ku is not needed.
53/ The most troublesome thing about Swahili verbs is the way they form the negative. There are two different problems connected with negatives. One is that the negative tenses don't correspond exactly to the affirmative tenses. There is only one negative to go with the na , a , and hu tenses (par. 29, 30, 31, 56). 54/ The negative of the me tense may be formed with -ja- or with -ku-, depending on the meaning. In this latter case, it is identical with the negative of the li tense.
55/ The other problem with negatives is that the learner should be prepared to find that each negative tense is formed in its own peculiar way. (It isn't quite that complicated, but if one starts with that assumption, then the similarities
267
