¹ The present study began as a term paper for a course conducted by Bernard Bloch in the spring of 1948. It owes a great deal to Bloch's criticisms and suggestions; the section on syntax (to follow) is modeled on his treatment of Japanese (Studies in colloquial Japanese II, Lg. 22.200–48 [1946]). I am also greatly indebted to Floyd G. Lounsbury for many helpful pointers. The method of positional analysis adopted here, by which morphemes are classified according to the order of their occurrence in a construction, derives from a paper read by Lounsbury before the Yale Linguistic Club in the summer of 1948. For a detailed exposition of this method see his dissertation, Comparative Iroquoian morphology (ms.), Yale University, 1949.

The study is based chiefly on material recorded by the class in Bloch's course from Gabriel Escobar's dictation, including one folktale of approximately 50 sentences (The fox, the mouse, and the tar doll). Subsequently, Lounsbury recorded another text of about the same length and put it at my disposal.

A useful list of Kechua affixes, with a vocabulary list, is found in J. M. B. Farfan, La clave del lenguaje quechua del Cusco (Lima, 1941). The author dismisses the matter of structure, however, with a single statement: ‘Los sufijos en el quechua pueden combinarse de mil diferentes maneras.‘ (References to this work will be identified by the author's name alone.)

² The phonemic analysis here presented was worked out by Bloch and Lounsbury.

³ Final [a] ends with slight nasalization.

⁴ See Charles F. Hockett, Problems of morphemic analysis, Lg. 23.321–43 (1947).

⁵ See Hockett. In the symbolic representation of morpheme sequences, an empty morhp will be denoted by Em.

⁶ See Hockett, op.cit. 322.

⁷ It may be that all noun roots can take both noun and verb suffixes, but the recorded material is not sufficient to prove this. For the present, this classification of roots has been adopted to distinguish noun stems that occur only with noun suffixes from those that sometimes occur also with verb suffixes.

⁸ A glance through Farfan shows that compounding is more complicated than this, sometimes involving several root forms or combinations of derived forms. Examples in this paper are limited to those found in the recorded material.

⁹ In the examples below, each complex form will usually be followed by a symbolic representation of the morpheme sequence involved. Where an example contains more than one word, the symbols refer only to the one word under discussion, hyphenated in the example.

¹⁰ Adverbs are words which form the center of an adverbial expression (see Syntax §7); particles are uninflected words of limited distribution which sometimes occur in combination with certain of the unrestricted suffixes (see §3.32 and §3.36).

¹¹ Numbers in parentheses are class numbers of the morphemes concerned, shown at the heads of the columns in Table 1. Where a class contains more than one morpheme, the particular morpheme is identified by a numeral after a decimal point.

¹² Hereafter, references to verb stems will also include verb bases, which are grammatically equivalent to stems.

¹³ On the basis of the complementation of rα- and ru-, one would expect the same distribution for pa-/pu- (2.3). But pu- occurs before the high vowel i (ruwa-pu-ηki ‘you will make‘).

¹⁴ The following abbreviations will be used to identify the pronominal suffixes :

I first person S singular PI plural inclusive

II second person P plural PE plural exclusive

III third person

— III non-third person

¹⁵ The II P morpheme -cis and the I PE morpheme -ku do not both occur in the same form.

¹⁶ Second person plural will be rendered in English (with apologies) as ‘you-all’.

¹⁷ For the distribution of these two future morphemes see §3.26.13–14, where they are treated in conjunction with the pronominal suffixes.

¹⁸ When asked for a form translating ‘would you-all’, the informant at first replied that there was none, but later offered the form here cited. It appears to be a combination of muna-wa-qh ‘you would like’ and the suffix conglomerate -cis-man-cu ‘could you?‘. In any case, it is a form of doubtful validity.

¹⁹ Note the concord: lloqhsi-qh-ta agrees with the form that it modifies, aya ucu-ta. For the tactical analysis, see Syntax §6.2.1.

²⁰ The verb tukuy ‘to pretend, to change’ is modified by a nominal adverb with the suffix maη ‘in, into’; e.g. qaqa-maη tukuη ‘he changes into a rock’, seryu seryu-maη tukuη ‘he pretended to be very serious’.

²¹ Em = empy morph. See §2.12 and fn. 5.

²² Since the morphs of Sets Two and Five are identical, it would be possible to treat them as a single set. They are separated here in order to complete the paradigms of verb suffixes and noun suffixes.

^22a [F. G. Lounsbury adds that future forms with second-person object have the endings I S F sa-y-ki ‘I shall ... you’, sa-y-ki-cis ‘I shall ... you-all’, III S F su-η-ki ‘he will ...you’, su-η-ki-cis ‘he will ... you-all’, I PE F sa-y-ki-ku ‘we (excl.) shall ... you or you-all’, III P F su-η-ki-ku ‘they will ... you or you-all’. These endings can be constructed from suffixes of Set Five, together with the choice of sa or su as indicated in Set Six. Examples: q'asu-sa-y-ki ‘I shall hit you’, qankunata q'asu-sa-y-ki-cis ‘I shall hit you-all’, q' asu-su-η-ki-cis ‘he will hit you-all’, q'asu-su-η-ki-ku ‘they will hit you or you-all’.—BB]

²³ A homophonous morpheme qa, occurring after nouns and pronouns, serves as a sentence connective or as an intensive. See §3.35 for a treatment of this morpheme.

²⁴ Farfan includes siki among his more than 30 suffixes. He says that it is suffixed to verbs and nouns and ‘denota inclinación: llullay.siki “propenso a la mentira”.‘

²⁵ It is not altogether clear whether kama is a suffix or a particle. It is tentatively classed among suffixes (a) because the stem which precedes it has no accent on the second-last syllable (yawar-ci-na-η-kama is stressed on ka, not on ci), and (b) because its distribution resembles that of paqh and some of the other adverbial suffixes.

^25a [F. G. Lounsbury points out that this morpheme is not an ‘unrestricted’ but a ‘restricted’ suffix, as is shown by its occurrence after the diminutive ca and before the goal suffix ta (e.g. tuti-ca-lla-manta and despasi-lla-ta in the examples below). Accordingly the morpheme must, be removed from Table 3 and inserted in Table 1, somewhere between column 11 and column 18; there is not enough evidence available to determine its exact position.—BB]

²⁶ Adv = adverb, not adverbial suffix (Ad); see fn. 10.

²⁷ For a definition of particles, see fn. 10.