Some grammatical aspects of Aymara will be discussed in this chapter; however, it is far from being a grammar of the Qoya language. The structural characteristics of Aymara will be discussed using a few examples of sentences of that language, so that readers who do not speak it will at least have an idea of the grammatical terminology which we are forced to use in our analysis of the logical structure of Aymara.

One of the main differences between Aymara and Spanish lies in the fact that Aymara sentences are not made up of words, but of strings of syllables linked together, i.e., a succession of suffixes added to a nucleus. The grammatical function and meaning of these strings correspond to several words in Spanish. For instance, the following string is a complete statement:

"manqayarapiskatapawa" ("You would have been having him eat it")

In this string, the nucleus is the verbal root "manqa", from the verb "manqaña" (to eat). The following suffixes have been added to this nucleus: "ya" which transforms the verb "to eat" into its homologue: "to make to eat"; "rapi", indicates that the verb action is to or upon somebody; "ska" indicates the present potential modality; "ta" is the ending for the second person singular ("you"); "pa" specifies that the recipient of the action is "him" (or "her"); and finally "wa" stresses it is a perfective statement, translated in this case by the auxiliary verb "to have."

Although in the example above, the string can be interpreted rather well in Spanish, some translation difficulties should be pointed out:

a) Although the enclitic pronominal expression "it... to him" (1) accurately translates the effect of the combination of suffixes "rapi" and "pa", there is no one-to-one correspondence between the Spanish pronouns and the Aymara suffixes. As a general rule, strings cannot be interpreted on the basis of a bi-unique correspondence between Spanish words and Aymara suffixes.

b) The modality indicated by the suffix "ska" is translated using the Spanish subjunctive, although in Aymara it is a present conditional. This is so because in Spanish the Potential is, fundamentally, a tense (future relative); when it is used to form a present conditional, one must use, the subjunctive, which has some connotation of "irreality", approaching, but not completely conveying the notion of "conditionality." Mood and tense are not always independent in Spanish; however, in Aymara an absolute distinction is made between tense and (logical) mood (2). Accordingly, for a great many expressions not all shades of meaning can be rendered in translation. For example:

cannot be differentiated in Spanish using the "Potential" (or "Conditional") tense when the suffix corresponding to the modality of possibility is added: In Aymara, the last sentence involves the Potential modality (possibility) in the Present Perfect tense; the Subjunctive mood has been used in the translation because Spanish does not have a "Preterite Conditional". These difficulties will be explained in greater detail throughout this book. They are mentioned here to bring to the reader's attention the fact that grammatical categories are not the same in Spanish and Aymara.

c) The controversial suffix "wa", a source of many misunderstandings, appears in the examples given above. Bertonio thought "wa" had only an ornamental function; however, he sometimes translated it as "is"; for instance, "jaça" ("big"), "jaçawa" ("is big"). Nowadays, "wa" is translated either as "be" or "have", depending upon the content. In Spanish, the auxiliary verb "to "have" is used to form a "perfect" tense, and implies that the action has been completed. To a certain degree this is equivalent to the emphasis added by the suffix "wa". However, this correspondence with the auxiliar verb "to have" has several limitations. First, in Spanish the auxiliar "to have" involves a change in verb tense, while "wa" does not. For example:

Second, "wa" can be attached to different elements in the sentence; thus, the notion of perfect action or of certainty about an event can be added to the predicate, the subject, or any other element. The impact of "wa" in different parts of the sentence can be shown in the following sentences, the analysis of which follows Nicolás Fernández Naranjo's interpretation (NFN 1): Several Spanish auxiliary verbs must be used to translatethe meaning of "wa" in these three sentences. In the first sentence "to have" has been used; however, in the second and third "it is...who" must be used to correctly show the stress added by this suffix when it is not a part of the predicate.

The suffix "wa" also has another connotation, which has no Spanish equivalent, but which is extremely important to understand the levels of logical modality in Aymara. As Martha Hardman (MHB1) pointed out, the Qoya language has a very peculiar grammatical categozy, which she calls the "postulate relative to the data source." Evidently, people who think in Aymara (even if they express themselves in Spanish) are used to specifying whether they are certain of what they say. The suffix "wa" indicates that what is said is reliable. For example, in the statement:

'Bolivar' aj markaru puriwa

The Bolivar has come to this town the suffix not only implies that Bolivar's arrival in town is coropleted, but also that the speaker is certain of it (because he witnessed the event). The suffix "wa" can be used to convey just the notion of a perfect action without implying the speaker can attest to the accuracy of this statement, as in the following example:

'Bolivar' aj ak markaru puri siwa

The Bolivar has come to this town, people say that additional element, "siwa", means literally "he/she has said"; in local colloquial Spanish it is "dice ché" or "diciendo" /people say/, and indicates that the speaker has only an indirect knowledge of that event.

The notions of perfect action and constancy indicated by "wa" should not be confused with the notion of certainty (modality of logical necessity) indicated by the suffix "pi". For example:

or also In this case, the added element "pues" /for sure/, pronounced "pss" in the variety of Spanish spoken in La Paz, indicates that in the mind of the speaker the statement is necessarily true, not because he knows it directly, but because he has reached this conclusion by logical inference. A perfect tense is not necessary in Spanish; the above examples might have been translated "él vino pues" /he came for sure/ or "él pss vino" /he for sure came/.

This preliminary discussion of the peculiarities of Aymara logic should suffice to show the reader that misunderstanding exists not only between Aymara- and Spanish-speaking people, but also between Aymara-thinking people (who speak Spanish) and Spanish-thinking people (even if they speak Aymara).

Although the analysis of misunderstanding is further complicated by the fact that some Aymara-thinking people speak Spanish, the study of the colloquial variety of Spanish spoken in the Altiplano is a useful tool for analyzing Aymara logic. In this regard, Tarifa's grammar (ETA) is a valuable contribution, because the numerous statements in Aymara recorded in it are translated using Bolivianisms.

The need felt by Aymara-thinking people to express the notion of certain knowledge has led to alterations in the syntax of Spanish, which does not have this grammatical category. These distortions have been thoroughly studied and explained by N. Fernandez Naranjo in his Dictionary of Bolivianisms (NFN). This author points out that "preference for the Present Perfect over the Preterite i.e., "has come" instead of "came" is due to the psychological need to express the certainty conveyed by the suffix "wa" in Aymara.

Another fine point of Aymara graromar should be brought to the reader's attention, viz, the existence of a "letter" which, is not pronounced and which causes the vowel preceding it to dis In this book this letter is referred to as an "elitor", and the symbol "/" is used to represent it. The elision can occur either in the middle or at the end of a string. Also, the elitor may appear alone, followed by other letters, or joined to some suffixes, and thus, it plays a very important role in the order of suffixes in any string. For example, the suffix "/ka", by itself forms the present modality in the conjugation:

The combination of suffixes "/ka" and "ti" generates a negation: In this example, the string "utjkiti" is more complex than it would appear at first glance. Its structure is:


Following are the symbols used in this formula: "." means "linked to", i.e., the continuous union of that which precedes and whatever follows the symbol; "RVB" means "verbal root", which in our example is "utja" (from the verb "utjaña", equivalent to "/there/ be", "be in existence"); "I2" is the second suffix in category I (Modal Inductor), which in this case is "/ka" , corresponds to the present modality (gerund); "T1", the first suffix in category T(Tense)correspondsto the Present-Preterite of Aymara, which is the elitor "/"; P4 is the fourth suffix in category P (Person) and corresponds to the fourth person ("he" or "she") of Aymara; and finally "L7", "ti", is the seventh suffix in category L (Logic), which corresponds to interrogative statements (when combined with I2 it generates negation). Thus, the string "utjkiti" is derived from:

RVB.I2.Tl.P4.L7. = utja./ka./.i.ti = utj/ka/iti = utjkiti

Although to a Spanish-thinking person this may look incredible, it is a very simple string which any Aymara 7-year old can master perfectly even if he cannot write!

In the above example, the adverb of negation "jani" (no) is followed by the suffix "wa" (I have direct knowledge that there is no bread); the elision of the last "a" is due to the addition of the suffix "K1", which also includes the letter "/". Below, the syntactic role of Kl, the first suffix in category K (Kasus) will be explained.

For the time being, the reader must bear in mind that the distinction between suffixes does not depend solely upon the particles ot letters which comprise them. Both the relative order of a suffix in any string and its function are also very important. This explains why "/" is sometimes classified as T1 and at other tiroes as K1.

The same applies to the particle "ni". This particle can represent up to four different suffixes, depending on its position and function. For example:

The Qoya language has more than twenty suffix categories; some of these include only two suffixes, and others more than ten, for a total of over two hundred suffixes, some of which may be combined within the same category (cf. Bertonio's "machinery of particles"), a fact which might discourage beginners trying to learn this language. Fortunately, any chance of confusion is eliminated because these suffix categories are clear-cut and their use is very consistent and easy.

This short monograph will analyze those suffixes directly involved in the logic of statements. The others, which will neccessarily appear in the examples, will not be explained systematically, but will only be translated into Spanish, with their meaning in Aymara rendered as accurately as possible.

In Aymara logic there are only nine simple suffixes which fall into two categories: I (Modal Inductor) and L (Logic). These suffixes will be analyzed in more detail in chapter 4, following a discussion of some essential concepts relating to trivalent logic and the use of truth-values, necessary to an understanding of the modalities which these suffixes generate. Before concluding this brief presentation of some special characteristics of Aymara, the author would like to comment upon two aspects which, though well known to scholars of Aymara, are not always clearly understood by "communicators" even if they speak (grammatically) correct Aymara.

In Spanish, the article plays two roles: 1) it limits and gives definiteness to nouns and, in some cases, to verbs (for example, "el tanto andar me cansó" / "so much walking made me tired"/; 2) it indicates the gender of the noun. In Aymara, nouns have no grammatically marked gender. Some nouns which refer to persons are specifically feminine or masculine, i.e., "warmi" (woman) and "çaça" (male, man), but, in general, things have no gender. For instance, "inti" (sun) is neither masculine like "el sol" in Spanish, nor feminine like "die Sonne" in German. The word "jaqe" means "man", in the general sense of "human being" (Mensch in German), but has no gender and can be used both for men and women. Also, the pronoun "jupa" means both "he" and "she". Thus, there can be no "sexism" when one speaks Aymara.

However, even if there are no articles in Aymara, there are, as is explained below, two suffixes, K1 and L1, which have greater syntactic flexibility than Spanish articles and may determine not only nouns, but also any complete string. For instance, in Aymara determiners can be used with proper names, whereas proper names in Spanish cannot be preceded by a definite article. For example:

'Pedroi puriwa.' (The Peter has arrived)

In this example, the subject consists of the string NOM.Ll.Kl, that is to say, the name plus the suffixes "ja" and "/", which determine it (implying: The little Peter, the one we know well). This feature is apparent in the colloquial Spansih spoken by Aymara-thinking people, who usually add the article and the diminutive ending to proper names, to make up for the missing functions of Ll and K1. Speakers feel that the Spanish sentence "Pedro ha llegado" /Peter has arrived/ is "bare" (the render may wonder which Peter it is).

Although the functions of the Spanish definite articles are covered by suffixes Kl and L1, they fall into the syntactic category of "pointers" as found in artificial computer languages. Like pointers in a data base, the string which contains the pointer is linked by K1 and L1 with w other string(s) containing determining or characterizing data.

The reader should be aware of another aspect: In Aymara there are 4 personal pronouns; in Spanish there are only three. The 4 Aymara pronouns are:

singular plural
1) jiwsa (you and me, both of us) jiwsanaka (all of us, inclusive term)
2) juma (you /sing/) jumanaka (you /plural/)
3) naja (I) nãnaka (we, mine, exclusive term)
4) jupa (he, she jupanak (they)

The "j" in "naja" is very soft; this word is sometimes pronounced "nä" (double vowel) and also "naya". The pronouns have been listed in their order of priority in Aymara. Conversation always involves a dialogue between two people (sometimes a few people), represented by "jiwsa", translated into colloquial Spanish by "nosotros dositos" /both of us/, or also "nosotritos" /those of us who are talking/. "jiwsa" is an archaic form (LBI, DTR); today the common form is "jiwasa".

To form plural pronouns; the plural suffix "naka" and a verb inflection are added to the singular pronouns; for example:

The suffix "pja" indicates plural in the conjugation. This is one of the few particles which has undergone some phonetic change: Ancient grammars (LB, DTR) record its pronunciation as "pesqa"; however, its position in the predicate string has remained unchanged.

The plural is formed by adding "naka" to nouns, but never to adjectives. This element is almost, but not quite, equivalent to the Spanish "-s". It is closer to the quantitative adjective "several". This causes some communication problems. When Spanish-thinking persons speak Aymara they use "naka" too often because of the frequency of the plural ending "-s" in Spanish. They mistakenly assume that gramatical categories of Spanish are "universally correct." For instance, the sentence:

Is translated into Aymara by Spanish-thinking people as follows: This "translation" sounds ridiculous and would be irritating to an Aymara. If the sentence were translated back into Spanish it would mean something like this: In correct Aymara " naka" is not required, and the adjective must always precede the noun: The excessive use of "naka" is very common among Spanish-thinking religious preachers.

Sometimes people erroneously assume that the repetition of a noun makes it plural. This is not so. In Aymara such a repetition generates a new collective (singular) noun. For example: "qala" (stone); "qala qala" (stony ground).

Finally, the terms "translation" and "explanation" as used in this monograph must be defined. Translation from one language into another involves generating sentences which closely represent the meaning of the sentences in the original language. It is not necessary for this faithful representation to be a word-for-word translation. What matters most is that the sentences in the target language accurately convey the meaning of the sentences in the source language.

Evaluating a translation is far from easy because it involves defining criteria to determine how faithful the sentence in the target language is to the idea in the source language. These criteria cannot only be syntactic and semantic. For instance, among the common people in La Paz one frequently hears , sentences like:

This sentence cannot be translated into English, although it may be grammatically represented by: Translation is difficult because in English the article not used before names, there are no diminutive endings, and there is no verb tense for the pseudo-Pluperfect of surprise. These grammatical categories have been transferred from Aymara into the variety of Spanish spoken in the Altiplano (the correct Pluperfect forro in Spanish is "habia", not "habiá").

Of course, an English-speaking person can be given an explanation of the real meaning of any sentence using illustrative examples, but this is not a translation, according to the definition given above.

This work does not attempt to discuss the many conceptual categories commonly used in Aymara. Instead, the subject of study is the logical categories implicit in the syntax of the Qoya language. As these categories sometimes do not have a Spanish equivalent, it will not always be possible to translate statements from Aymara into Spanish; however, they will be explained as precisely as is required by logic using some mathematical symbols. Therefore, the reader must be patient with and tolerant of our utilization of certain symbols and truth-tables, as these tables will help to understand more clearly the differences between Spanish and Aymara logic, and will also help to avoid imprecise wordiness.

Note: The symbols used to transcribe Aymara phonemes are explained in Appendix D.

(1) Translator's Note: It is not enclitic in this example (neither in Spanish nor in English)

(2) Translator's Note: IGR says "modality."

Chapter index