¹ Both -a and -ε may be followed only by the suffix -ni of the second person plural or by certain enclitic elements.

² Cf. §30.

³ -i seems to have no morphological significance here, but serves merely as a vocalic complement to the consonantal stem. Cf. §4.

⁴ Cf. §4.

⁵ Cf. §3, b.

⁶ This is not a true vowel. It could just as well be written as y.

^6a That is, require ε rather than i.

⁷ Cf. Edward Sapir, Language 106–7; New York 1921.

⁸ It all depends upon one's interpretation of certain auxiliary elements. Cf. §37. Absolute demarcation between these categories is very nearly impossible.

⁹ Or A + B .. .

¹⁰ This is in no sense a unique characteristic.

¹¹ Cf. §37, j.

¹² Cf. §29.

¹³ Cf. §30.

¹⁴ Cf. §33.

¹⁵ Cf. §25.

¹⁶ The dots indicate separation. -ku- is prefixed next to the stem and -ni is suffixed as the final element in the word. Example: nakugwírani ‘I have caught you’.

¹⁷ w is here undoubtedly an old concordance, u-, which has been preserved in this tense. It occurs in other places; e. g., munt'u u·yu ‘this person’, u- is the regular concordance for this class in Isizulu.

¹⁸ This form clearly shows that there is no suffix -ama existing as a live form expressive of the stative in this language, as is said to be the case in some of the other members of the Bantu family.

¹⁹ Cf. §30, A: d-2.

^19a I.e., the plural as polite form of address to a single individual.

²⁰ Cf. §25.

²¹ This is not a concordance. It may be called an expletive prefix, since its function seems to be entirely of that nature.

²² Cf. §30.

²³ J. Torrend, S. J., A Comparative Grammar of the South-African Bantu Languages 222; London 1891.

²⁴ Cf. §28.

²⁵ Allowing ‘might’ to stand for the volitional phase and its frustration.

²⁶ This has been called the ‘dependent mood’, but, as we shall see, the terminology is not very satisfactory. See Clement M. Doke, Text Book of Zulu Grammar 172; Johannesburg 1927.

²⁷ A piece of cloth which is worn so as to hang down apron-like behind.

²⁸ A type of dance.

²⁹ Carl Meinhof, Grundzüge einer vergleichenden Grammatik der Bantusprachen 64; Berlin 1906.

³⁰ This form of imperative seems to be less emphatic or somewhat more mild than that in which a substantival or independent pronominal object is involved. That the mood may be changed by addition of the objective concord is not in the least far-fetched. In Spanish, e. g., there is an instance to some extent parallel, wherein introduction of a negative in a command causes the verb to take the subjunctive form: Ponlo en la mesa. But No lo pongas en la mesa. No escribas eso, Inés.

As we shall see, the suffix-ε is more appropriate for a formal reference than for a strong, more imperious one.

³¹ In the first instance the speaker denies that which is expected or contingent, as in He has not come—yet, and the second is a denial of that which is at the time of speaking true, as in He had not come—then. This, again, may be compared to such Spanish forms as Niego que lo haya hecho, although there are other factors, que and change of subject, involved. Subjunctives also occur in negation in the Tsimshian (Nass River dialect) language. See Franz Boas (ed.), Handbook of American Indian Languages, Part 1.403; Washington 1911.

³² This terminology has been taken from Charles K. Ogden and Ivor A. Richards, The Meaning of Meaning, Chapter 10; London 1923.

³³ Handbook of American Indian Languages, Part 1,399.

³⁴ Ibid. 495–6.

³⁵ It is very difficult to give a definition of voice on a purely functional basis, such as would be of universal application. It is closely related to the question of relationship among subject, verb, and predicate, which is largely a matter of formalism. The linguist's recognition of certain voices in a language (except active and passive) is often based primarily on formal analogy.

³⁶ Cf. §26.

³⁷ Cf. §25.

³⁸ From the Chewa point of view, a woman does not marry, but is married by a man: kukwáti·wa ‘to become married by a man’.

³⁹ kↄ·'di merely indicates that that which follows is a question. Cf. §37, k: 6 and §53.

⁴⁰ See Charles K. Ogden and Ivor A. Richards, The Meaning of Meaning 371–81 ; London 1923. Cf. also §33.

⁴¹ Cf. §28.

⁴² Clement M. Doke, Text Book of Zulu Grammar 238; Johannesburg 1927.

⁴³ Carl Meinhof, Grundzüge einer vergleichenden Grammatik der Bantusprachen 81; Berlin 1906.

⁴⁴ See Alice Werner, Introductory Sketch of the Bantu Languages 186–98; London 1919.

⁴⁵ In such instances as these, Doke regards the verb as derived from the radical, and apparently never the reverse. See op. cit. 147–8, 238–52. This seems to be the usual opinion. Cf. Werner loc. cit. I should hesitate in pronouncing either form the older in Chichewa. It is possible that derivation may proceed in either direction, although there is perhaps more a priori justification for assuming the radical descriptive to be the older.

⁴⁶ Such a syllable also has considerably more ‘weight’ or force than that which normally appears.

⁴⁷ See Carl Meinhof, Grundzüge einer vergleichenden Grammatik der Bantusprachen 81; Berlin 1906, and Alice Werner, Introductory Sketch of the Bantu Languages 197–98; London 1919.

⁴⁸ Cf. §37, k: 4.

⁴⁹ Cf. §29, e.

⁵⁰ ‘A word-group is a combination or fusion of two or more words brought about through vowel elision (or coalescence).‘ Clement M. Doke, The Phonetics of the Zulu Language 190; Johannesburg 1926.

⁵¹ A high pitch on the last syllable in this case will be determined by the word which follows in the ‘sense-group’. Cf. Doke, ibid. 194.

⁵² Cf. §30. A: f.

⁵³ Cf. §29, d.

⁵⁴ See j, below.

⁵⁵ See §30, A: h and B: c.

⁵⁶ Cf. §43.

⁵⁷ Clement M. Doke, Text Book of Zulu Grammar 180; Johannesburg 1927.

⁵⁸ Cf. c, above.

⁵⁹ ‘The word is merely a form, a definitely molded entity that takes in as much or as little of the conceptual material of the whole thought as the genius of the language cares to allow.‘ Edward Sapir, Language 33; New York 1921.

⁶⁰ Cf. §36, c.