NAVAJO SYLLOGISMS: STRUCTURES AND USE 1. Abstract

Transcription

1 3 NAVAJO SYLLOGISMS: STRUCTURES AND USE Abstract James F. Hamill Miami University Oxford, Ohio Modern cognitive anthropology has shifted its interests from analyzing groups of related lexemes to analyzing discourse with emphasis on the schemata that provide coherence to speech. Logical inference patterns such as syllogisms play an important role in these schemata and have received attention in the recent literature. The results of an investigation into the structural validity and folkloric use of Navajo syllogisms are reported in this article. The syllogism is found to be a valid reasoning pattern in Navajo and a key schematic feature of Navajo folk tales. INTRODUCTION In the mid and late 0s, cognitive anthropology concentrated on analyzing groups of related lexemes. Folk taxonomies and paradigms of kin-terms appeared with alarming regularity in the literature, but those appearances have moderated. Modern cognitive anthropologists have shifted their interest away from isolated words and toward discourse. This shift is a change in emphasis, rather than theory; the assumption remains that language is "the prime inroad to understanding group life." [Agar 8:8]. The study of discourse requires new analytical tools, and one of the most important of these tools is the "schema," the formal structure which makes discourse coherent. This paper concerns logic, which is one of the many aspects of schema. The literature shows that theories in both linguistics and anthropology assume that humans reason logically. Both cognitive psychologists and linguistic anthropologists have produced empirical evidence of these theories in studies of syllogistic reasoning among widely dispersed peoples. The work reported here continues the investigation of syllogisms among the Navajo. It establishes the validity of Navajo syllogisms and demonstrates their use as part of the "grammar" of Navajo folklore. The work was conducted on the Navajo reservation in the summers of 77, 78, and 80. Eight native speakers of Navajo, ranging in age from nine to seventy years, participated in the study. Four were monolingual in Navajo, and four were bilingual in Navajo and English. BACKGROUND While there is a need to investigate more powerful reasoning patterns, such as propositional logics, in this work I continued to concentrate on syllogisms because they represent a relatively straightforward logic that is

2 amenable to both paradigmatic and textual analysis. The syllogistic argument contains three statements that deal with class-inclusion relationships. The argument must be constructed so that the first two sentences in the argument force the third sentence as a conclusion. There are four kinds of class-inclusion propositions that make up the syllogisms. They are traditionally labelled A, E, I, and O. A: All S is P. E: No S is P. I: Some S is P. 0: Some S is not P. The three statements in the syllogism are the conclusion, the major premise which mentions the predicate term of the conclusion, and the minor premise which mentions the subject term of the conclusion [Barker :], The terms mention classes of objects as in a folk taxonomy. There are different ways to combine these four proposition types in groups of three ( major premise x minor premise x conclusion = ). These combinations are called moods. The first two statements must "share" one term, term M M," the middle term. There are four different ways in which that term can appear in the first two statements. The four arrangements are called figures. Combining figure and mood, there are possible syllogisms ( mood x figure = ), but only are valid according to the rules of philosophical logic (see Table ). If only the argument (major premise and minor premise) is considered, there are possible syllogisms ( major premise x minor premise x figure = ). Table. Valid Syllogisms [after Reichenbach 7;0. FIGURE i II III IV Major Premise Minor Premise MS RM SM RM MS MR SM MR MOOD Valid in all Figures EAO EIO * Valid in 3 Figures AAI Valid in Figures EAE ATT AEE AEO IAI * Valid in Figure AAA AOO Aw 0A0 + Legend: M«shared (middle) term; S«conclusion predicate term; R*conclusion subject term. HUMAN REASONING IN LINGUISTIC AND ANTHROPOLOGICAL THEORY Linguistic and anthropological theories assume that humans are capable of, and do use, logic in their thought processes. All modern theories

3 of discourse rely heavily on an assumed human capacity to reason logically. Grice's influential work proposes a n conversational implicature" as being essentially connected with certain general features of discourse. These "implicatures" are much like "if... then..." statements in a propositional calculus, except they are constrained in human conversations by a "cooperative principle" that limits the "Quality," "Quantity," "Relation," and "Manner" of what people say in conversation [7:]. Anglo-American theories of discourse are not the only ones that rely on an assumed capacity to reason logically. Van Dijk, the Dutch linguist, views a theory of discourse as a "text grammar," which is a model "for the instructive knowledge (competence) of native speakers to produce/interpret coherent texts of the language." To him, the underlying structure of these texts, and therefore both the text grammar and the mind which knows the grammar, are similar to the "internal structure of the propositions in a model predicate logic" [7:307]. Clearly both Anglo-American and European linguistic theories view logical reasoning as an artifact of being human. Most modern anthropologists also view the ability to reason logically as an innate human capacity which may not be usable to distinguish one culture from another, but which can be used in explanations of cultural universals. Even Melville Herskovits, the father of cultural relativism, asserts that logical thinking is innately human. He relates an experience in Africa where he studied a certain group of people who believed that the younger child is "sharper of wit than his older siblings." These people had argued that the children resembled their parents, and, since, as men grow older, they grow in wisdom and experience, the youngest child of a marriage should possess "a surer, more alert awareness." Given these premises, Herskovits claims, "the logic is impeccable," and that cultural relativism could claim only variation in premises, not variation in thought process [Herskovits :3-3]. Claude Le*vi-Strauss correctly defines logic as "the establishment of necessary conclusions" [:3], and goes on to state that "the savage mind is logical in the same sense and in the same fashion as ours..." [:8]. Elsewhere he asserts "...the kind of logic in mythical thought is as rigorous as that of modern science and the difference lies, not in the quality of the intellectual process, but in the nature of the things to which it is applied" [7:7]. Both Le*vi-Strauss and Herskovits must assert that all humans have the ability to reason logically in order for their theories to make sense. Herskovits needs a universal human logicality to disspell the notion of "simple culture" [:30] and Le*vi-Strauss needs a logical human in order to account for "the logic of totemic classifications" [:33-7] and social organization [7] and myth [70]. While we take it as given that humans possess some capacity to reason logically, this assumption does not mean that we know the form and structure that such capacity takes in any particular language/culture setting. That information can only come from field work. Fortunately the question interests both cognitive psychologists and linguistic anthropologists, who have produced substantial research to the subject.

4 Most research in ethno-logic has concentrated on syllogisms. In West Africa, Michael Cole and his colleagues found valid syllogistic reasoning when the subject matter of the arguments concerned daily life [Cole et al. 7], Sylvia Scribner, also in West Africa, continued Cole's work and confirmed his conclusions [7:3-7], Similarly, in research among Siberian peasants, Luria found that the validity of syllogisms is an artifact of the abstractness of the premises [7:-7]. Edwin Hutchens [80] described valid syllogistic reasoning as a key factor in Trobriand land tenure disputes. The conclusion which these psychologists draw are consistent, both with one another, and with my own conclusions from the linguistic anthropology tradition [Hamill 78:-], THE VALIDITY OF NAVAJO SYLLOGISMS Folk taxonomies were used to test the validity of syllogisms in Navajo. The first step was to elicit the taxonomy given in Table. This was elicited from three consultants. Robert was a young man who was bilingual and biliterate in Navajo and English; Chee, an older man (over 0 years), was monolingual in Navajo, and the third, Annie, Chee's wife, was also monolingual. The taxonomy was elicited from each independently, as were their responses to the subsequent syllogism tests. Table. Syllogism Text Taxonomy, Nanlsi (plant) (A) tsin (tree) (B) " " ch'il (grass) (C) gad- ^^.t'iis ch'il diilghesii^^ ^^ch'l 'abe'e (cedar) (D) (cottonwood) (E) (snakeweed) (F) (milkweed) (G) After the taxonomy was elicited, each person was asked to supply the sentence frames that correspond to the A, E, I, and O sentences in Navajo. That is, they were asked, "How do you say 'All A T s are B T s (No A T s are B T s, etc.)* in Navajo?" This gave four sentence frames from which the sentences in Table 3 could be constructed. These sentences were then combined to form the syllogistic arguments (major and minor premises) that represent all of the possible syllogistic argument figures and moods. 3 Table 3. Sentence Frames.. All D is B. 7. No D is C. 3. Some B is not C.. Some D is A.. All B is A. 8. No B is C.. Some A is not C. 0. Some C is not B. 3. All D is A.. No C is B.. Some D is not C.. Some D is B.. Some B is D. 0. Some B is not D.. No C is D.. Some A is B.. Some A is not D. 7. Some B is A.. Some A is D.. Some A is not B. 8. Some C is not D. Legend: A, B, C, D, etc., refer to the taxa in Table. Thus, sequentially, "All D is B" would be glossed "All cedars are trees." ^

5 7 Once the possible arguments were constructed, each consultant was given each argument and asked to draw a conclusion. The conclusions proffered were either "correct" or "incorrect," (i.e., the conclusion was validly or invalidly drawn in terms of philosophical logic). In some cases (i.e., the A A syllogism of Figure II), drawing no conclusion is the "correct" conclusion. Validly drawn conclusions were noted. An improper conclusion could take one of two forms and was challenged accordingly. If the consultant gave no conclusion where he/she could have, he/she was asked if he/she agreed with a given conclusion. If a "wrong" conclusion was given, I asked him/her to argue toward that conclusion. The presentation of a valid argument to that conclusion was noted and considered a valid syllogism. The results of the paradigmatic tests showed that of the valid syllogisms recognized in philosophical logic, a total of were drawn in Navajo. Of the invalid syllogisms in philosophical logic, the Navajo consultants rejected. There were four conclusions that were apparently invalidly drawn and five that were apparently invalidly refused. I will first discuss those conclusions that were invalidly drawn. In several cases, consultants appeared to draw invalid conclusions. However, on further investigation, these conclusions were found to result from changes in premises that allowed the conclusions to be validly drawn. For example, when given the EO argument of Figure II: No grasses are trees Some plants are not trees Robert gave the conclusion "some plants are not grasses." When challenged on this conclusion he changed the minor premise to "some plants are trees." His conclusion follows. In all of the cases of invalidly drawn conclusions, similar changes in premises occurred. Furthermore, in all of these instances the change in premises was such that the proffered conclusion followed. Therefore no conclusion was invalidly drawn. In several cases conclusions were apparently invalidly refused. In these syllogisms it was possible to draw a conclusion, but consultants would not draw it and refused to agree with it when it was given to them. They believed the conclusions to be true but disagreed that it followed from the argument. In each of these cases the conclusion was a particular statement (Some A is B; Some A is not B) in which the corresponding universal statement (All A is B; No A is B, respectively) is true. All of the invalidly rejected conclusions are listed in Table. Table. Mood AAI EIO All EIO EIO Invalidly Rejected Conclusions, Figure I I III III IV Coneluaions Rejected (Gloss) Some cedars are plants Some trees are not grasses Some cedars are plants Some cedars are not grasses Some cedars are not grasses

6 8 The invalidly refused conclusions are interesting not only for their similarity of structure, but also because they serve to constrain the class of drawable conclusions. Because fewer conclusions can be drawn in the natural logic from the same structures, the logic must be seen as stricter than the formal logic proposed by philosophical logicians. This analysis contradicts the common view that natural logic (common sense?) is looser than rigorous self-conscious thought. SYLLOGISMS IN NAVAJO DISCOURSE To determine if syllogisms are used and salient in natural Navajo discourse, I analyzed one Navajo folktale for the syllogistic structures it contained. The same tale was collected from five individuals in three settings (one version was collected in the home of the consultant, three were collected in a school situation, and one in my home). Alfred, the principle consultant, was a 7-year-old monolingual speaker of Navajo. Three other consultants, Johnny, Ben and Thomas, were bilingual Navajo children (Navajo being the first language) who ranged in age from 8 to years. The last, Nancy, was a 3-year-old woman, bilingual and biliterate in Navajo and English. The analysis considered four different versions of this tale: one from Alfred, a second from Nancy, a third from Thomas, and a fourth version resulting from collaboration between Johnny and Ben. It is important to explain the substantial difference in content and length between the stories told by the three Navajo children and Alfred's story. I gathered these stories in one of my first field experiences. I had arranged to spend a summer on the Navajo reservation in order to study the logical content of Navajo folklore. As part of that arrangement, I taught a class of first through third grade children who were having problems with English. I had never taught a class below the college freshman level, so I had to deal with that adjustment as well as the culture shock that accompanies ethnographic research. Understandably, I made some mistakes. One of the most serious mistakes was asking the students in the class to write a coyote story. I was looking for ways to have the students write, and felt that if they were familiar with a topic, they could write about it more easily. I also knew that coyote stories were popular on the reservation, so I reasoned that an assignment to write a coyote story would be both enjoyable and beneficial. I did not realize that there was a serious taboo against telling coyote stories in the summer. Coyote stories are reserved for the winter. From the students' point of view, my assignment presented three problems. First, they had to deal with me, and I was exhibiting all of the negative behavior that reflects culture shock. Second, they had to write, which was a difficult task for them. If they had been good writers, they would not have been in the class. Finally, in order to perform that task they had to violate an important cultural restriction. Therefore, the brevity of the last two responses, Thomas's response and that of Johnny and Ben, was only to be expected.

7 Alfred, who was not in the class, tried to warn me about the taboo a few days earlier when I had collected the story from him. But, because I did not speak Navajo and because Nancy, my interpreter, did not translate it, I did not understand his message. I discovered later that he had agreed to tell me the story, in spite of my ignorance, because he felt that my interests were genuine. His family was also in need of the fee I paid him for telling me the story; one of his granddaughters was approaching puberty and required a kinaalda, a long and expensive ritual. Alfred*s status as a grandfather was important here because in that role he told stories. He was therefore familiar with the stories, living with them all of his life, and he was experienced in their telling. Indeed, he had a reputation in his local area for being a skilled orator. Because of this skill and knowledge, of the four his version most closely approaches the Navajo ideal of what the story is and how it should be told. I collected the story in Alfred*s home, with Nancy acting as interpreter. We recorded the story on audio tape, and later Ms. Martha Austin, now of Kayenta Research Associates, Inc., provided a transcription and word-for-word translation. Once the story was collected, Nancy and I went to my home and listened to the tape. It was at this time that Nancy gave me her version of the story. The name of the tale was "Coyote and Skunk,* one of a class of traditional folk tales known as Coyote Stories. This particular story seems to have a long history: Alfred said that his grandfather told him the story when he was a young boy. The main character in these stories is Coyote, a being of great mystical power, who was one of the first created. Coyote makes mischief for himself in all these stories, and this tale does not vary from that general pattern. The analysis of the Coyote story shows important use of syllogistic logic that is salient to native speakers of Navajo. In this story, Coyote, through a rather complicated series of actions, meets with his cousin Skunk. They conspire to trick and eat some prairie dogs and jack rabbits who live nearby. The ruse is successfully completed, and Coyote and Skunk kill many of the small animals. After they prepare the kill for consumption, Coyote attempts to trick Skunk so he can eat the largest share. However, Skunk sees through Coyote*s deception and is himself able to trick Coyote and get the lion*s share. Coyote, upon seeing that he has been tricked, pleads with Skunk for some of the kill, but Skunk takes advantage of Coyote*s greed and causes him to kill himself. Only Alfred*s version contained the entire story. However, all three mentioned the sequence that brings Coyote and Skunk together. Therefore I analyzed this portion of the story for the syllogistic logic it contained. The relevant portions are given in version A below, in a free English translation. Each sentence is numbered and given in the order it appeared in the original text.

8 0 Version A (Alfred) (Free English) Al. Coyote is hot and may be dying of thirst. A. He spoke thus, "I wish a cloud would be for my benefit." A3. Finally, as the day proceeded, the cloud came into being as he wished. A. Then he wished for a light sprinkle of rain to fall on him. A. Just as he wished a sprinkle fell on him as he walked. A. Thus it happens just as he says. A7. You see he spoke in this way, "I wish water would rise to make the soles of : feet moist as I walk along." A8. He said, "And I wish the water would rise to reach the middle of my arms as walk along." A. Thus they say he walked like that. A0. There is a wide area where many prairie dogs live. All. And they say he said, "I wish I would float where many prairie dogs live." A. Thus it happened like that. A3. He spoke out his wish thus, "I wish I would land where many prairie do live." A. And the water started to rise taking him with it and he landed where the wat receded and many prairie dogs live. A. He said, "I wish to stop floating where Skunk comes to drink." A. Thus he landed where the foam formed around the greasewood as he had wished. In A a wish is expressed and A3 its granting is indicated; a simil structure appears in A and A. We can symbolize that structure in tl following ways: Let z «Coyote's wish w «happen VI - that a cloud be for my benefit V - that a light sprinkle would fall on me

9 A would be premise (PI) and A3 would be conclusion (Cl): PI... VI is z (that a cloud be is Coyote's wish) Cl... V is w (a cloud happens) A would be premise (P) and A would be conclusion (C): P... V is z (a sprinkle of rainfall is Coyote's wish) C... V is w (a sprinkle happens) A does not function like A, A3, A or A; however it does have its logical place. It is the major premise. The storyteller tells us that it happens just as Coyote says it will. The logical structure of A is expressed in premise (P3): P3... All z is w (all of Coyote's wishes happen) Cl follows from P3 and PI, just as C follows from P3 and P. This is shown in syllogisms (SI) and (S): S.P3... All z is w (all of Coyote's wishes happen) PI... VI Is z (that a cloud be Coyote's wish) Cl... VI w (a cloud appears) S.P3... All z is w (all of Coyote's wishes happen) P... V is (a sprinkle of rain is Coyote's wish) C... V w (a sprinkle happens) After the major premise is stated, four more minor premises /conclusion (wish-granting) structures are encountered. A7 and A8 each state a wish. A states that the wish is granted and concludes the syllogisms. The structure is repeated in All, A, A3, A, A, and A. Indeed, every sentence in this excerpt, with the exception of Al and A0, serves an important logical function. The three remaining versions (B, C, and D) are not as complete as Alfred*s. This means that in relation to version A, the other three storytellers deleted some items. Their versions are given below. Version B, like version A, does not represent the entire text, but only the wish sequence. On the other hand, versions C and D are the complete texts of Thomas's, and Johnny's and Ben's responses, respectively. Version B (Nancy) Bl. Coyote was dying of heat and thirst. B. The more he thought about it he said to himself that he wished it would just sprinkle.

10 B3. After he wished it, it started to sprinkle and Coyote was very happy. B. He wished again that it would rain more than a sprinkle. B. He wished that he would walk in a running river. B. His wishes always came true when he wished it. B7. He wished that he would float in the river and land where the prairie dogs and rabbits live. B8. And after landing there, he wished again. B. He asked his cousin Skunk to discover him. BIO. When cousin Skunk came to get water he saw cousin Coyote. Version C (Thomas) (Thomas wrote this as one sentence but for purposes of analysis I have broken it up into three.) Cl. One day Coyote says, "I wish I get some rain." C. So I will wash to prairie dog house. C3. So I can get to eat. Version D (Johnny and Ben) Dl. One day the Coyote was walk down the rivers. It was too hoty [sic.]. D. There he said, "Rain come down and I'll be cool." D3. There was a skunk trying to kill the prairie dog. The coyote said, "I'll just pretending to lay down, then you just tell the prairie dog and rabbit that I am asleep." D. Then the skunk, rabbit and prairie dog was dancing around the Coyote. D. But the Coyote wasn't sleeping, he was just pretending to sleep and he had a stick in his hand and then having got up and kill the rabbit and the prairie dog. In none of these versions is every episode of Alfred's version mentioned, but that fact alone does not indicate much because versions C and D are severely attenuated. More importantly, no version mentions any episode that Alfred does not mention. That is, every sentence in versions B, C, and D is analogous to some sentence in version A. Table lists each sentence in version A, along with its logical function and analogues in versions B, C, and D.

11 3 Table, Logical Structures and Analogues in Four Versions of "Coyote and Skunk." Version A Versions B, C, and D Analogues Sentence Logical Function context minor premise conclusion minor premise conclusion major premise minor premise minor premise conclusion context minor premise conclusion minor premise conclusion minor premise conclusion Analogue elsewhere in the story 8 t, 0 3 3,, The major premises (All of Coyote's wishes come true) and the conclusions (The thing Coyote wished actually happened) in each of the syllogisms in this portion of the story are the same. These syllogisms are not essential to the story, which could have just as easily listed the events that brought Coyote and Skunk together. That is, the story could have gone something like: "One day Coyote was hot as he walked along and a cloud formed and cooled him. Then it began to sprinkle and soon it was raining very hard." Note that the minor premises are arranged in developmental sequence (Coyote wishes for a cloud, and then a sprinkle of rain, and then a hard rain and so on). Thus the content of the story is carried in the minor premises. We can account for the presence of these syllogisms by looking at these processes of story telling. The oral poet does not recite from memory, rather the story/poem is composed on the spot. If memory was the media of oral literature, then each instance of telling a story in a particular tradition would be the same, except for minor mistakes. But that is not the case not only do we find wide variation among storytellers we also find that a single poet will perform a story differently on different occasions. The teller actually uses a sort of grammar of the story at the core of which is the "formula" [Lord 0:30-7]. The formula, as defined by Perry, is w a group of words which is regularly employed under the same metrical conditions to express a given essential idea" [30:80]. The formulaic nature of Alfred's version is obvious. The formulas appear not

12 only in repeated groups of words, but also in repeated logico-linguistic structures (schema), in this case syllogisms. One may argue here that the syllogisms appear only in version A and not in the other three. Thus, even if they are emphasized, the emphasis is not getting across to less well-trained storytellers. The distribution of analogues in versions B, C, and D belie that idea. Except for Bl, B, and Dl, all of the sentences in versions B, C, and D are minor premises, as shown by the fact that they are expressed as Coyote's words, and not, for instance, a list of the events that brought Coyote and Skunk together. Johnny and Ben tell us that Coyote said, "Rain come and and Til be cool." According to Thomas, Coyote's words were, "I wish I get some rain...," and Nancy says that Coyote wished "that he would walk in a running river." All of these people are Navajo, and as Navajo they know who Coyote is, and they are aware of his powers. They know that all of Coyote's wishes come true (the major premise), and that things that Coyote wishes actually do happen (conclusion). Since this is general knowledge that the participants bring to the story, the information can be deleted and the story will suffer little or no loss of content. The fact that the wishes are viewed as Coyote's words, and not as events, indicates that these wishes are part of the underlying syllogistic structure that version A most fully expresses. This emphasis of logical structure is clearly an indication that the structure itself is part of the story's formula. CONCLUSIONS From this work we can draw two kinds of conclusions. One compares philosophical logic as it is found in textbooks to the structure information elicited in the paradigmatic tests. The other concerns the use of syllogistic reasoning in relatively natural Navajo discourse. In classical philosophical logic four rules account for the classes of acceptable and unacceptable syllogisms [Barker :]: Rule : In a valid syllogism, the middle term is extended at least once. Rule : In any valid syllogism, no term is extended in the conclusion that is not extended in the premises. Rule 3: No valid syllogism has two negative premises. Rule : Any valid syllogism has one negative premise if and only if it has a negative conclusion. When applied to all of the possible syllogisms, these rules generate the valid forms given in Table. However, these rules are not powerful enough to account for the natural patterns detailed above. They generate too many valid forms, and their explanatory value is weak in that they account for nothing other than the syllogism. The results of the paradigmatic tests call for a more powerful logic that disallows conclusions of the form

13 "some A is B" and "some A is not B" when the corresponding universal statements ("All AisB;" "No A is B") are true. I have argued elsewhere that we can account for these anomalies on the basis of taxonomic semantics [Hamill 7:8-]. In classical logic, the statement "some A is B" means that there exists an A that is a B, thus allowing the conclusions that my consultants rejected. But in a taxonomic sense, to say that "some cedars are trees" is misleading at best. To include semantics in our view of ethno-logic would take us a long way toward understanding human reasoning. The analysis of the story "Coyote and Skunk" approaches the question of use. It shows that in its complete version told by a professional storyteller logical structures comprise an important part of the "grammar" of the story. These logical structures are analogues of Perry's formulas that enable the oral poet to compose the story on the spot. Furthermore, the structure is salient to Navajo children; it stands out to them. Their use of Coyote*s words in their versions of the story clearly shows this understanding. Thus, this work demonstrates that syllogistic reasoning is valid to Navajo people and is used in their daily lives. It also adds empirical support for our assumptions about human logicality. Further work should investigate other logics, such as propositional reasoning, as well as the question of whether the logic is really a part of daily natural discourse. NOTES. The research that led to this paper was supported in part from a grant awarded by the Miami University Faculty Research Committee and funds from the College of Arts and Science of Miami University. I gratefully acknowledge their assistance. I also acknowledge the help of Ms. Martha Austin, who gave me a detailed word-for-word translation of the Coyote story. Dr. Oswald Werner, Mary Helen Creamer and all of the people on the Navajo Reservation who helped me in my work deserve my deepest gratitude. I would also like to thank Dr. George Esber, Dr. George Fathauer, Dr. Ann Davis, Mr. Jerome Perk, Ms. Deborah Hughes (all of Miami University), and several anonymous reviewers for their comments and suggestions that improved this paper. All remaining errors are of course my own.. All names are fictitious. 3. The arguments given to consultants are listed below. The numbers (,, 3, etc.) in Che list refer to the sentence on Table 3. Thus the AA syllogism of Figure I is listed on the protocol as: The English gloss of that argument would be: "All trees are plants"/"all cedars are trees" to which either sentence 3 or ("All cedars are plants" or "Some cedars are plants") would be a valid response.

14 Chart. Protocal Sentence Combinations. Figure II III IV Figure II III IV MOOD AA 3, 3 3 AI Jf AE 7, 8 7, AO IA JL 7 II IE loo 0 0 EA 8 7, 7, 7 _ 3 JL El 7 3 _ 8 J» Jf OA 0 _ 0 JL Jf OE 0 0 _8 0 _ It may be argued there that the granting of a wish is not logical; however, that misses the point of logic. A syllogism is a syllogism because of its formal structure, not because of its semantic content. Syllogistic rules mention only the relationship that must exist between premises in order for a conclusion to follow. To assert that wishes are not granted in the real world does not invalidate the logical structures in this story. It attacks the world view of the Navajos, but it does not attack the logic. REFERENCES CITED Agar, Michael H. 8 "Whatever Happened to Cognitive Anthropology: A Partial Review." HUMAN ORGANIZATION (l):8-8.

15 7 Barker, S. F. ELEMENTS OF LOGIC. New York: McGraw-Hill. Cole, M., J. Gay, J. Glick and D. Sharp 7 THE CULTURAL CONTET OF LEARNING AND THINKING: AN EPLORATION IN EPERIMENTAL AN- THROPOLOGY. New York: Basic Books. Grice, Paul 7 "Logic and Conversation." In SYNTA AND SEMANTICS. Peter Cole and Jerry Morgan, eds., pp. -8. New York: Academic Press. Hamill, J. F. 78 "Trans-Cultural Logic: Same or Different in Three Languages." In DISCOURSE AND IN- FERENCE IN COGNITIVE ANTHROPOLOGY. M. D. Loflin and R. SilverbeTg, eds., pp. -. The Hague: Mouton. 7 "Syllogistic Reasoning and Taxonomic Semantics." JOURNAL OF ANTHROPOLOGICAL RESEARCH 3()-.8-. Herskovits, M. CULTURAL ANTHROPOLOGY. New York: Alfred M. Knopf. Hutchens, E. 80 CULTURE AND INFERENCE: A TROBRIAND CASE STUDY. Cambridge: Harvard University Press. Levi-Strauss, Claude THE SAVAGE MIND. Chicago: University of Chicago Press. 7 STRUCTURAL ANTHROPOLOGY. Garden City: Anchor Books. 70 THE RAW AND THE COOKED. New York: Harper Torch Books. Lord, Albert 0 THE SINGER OF TALES. Cambridge: Harvard University Press. Luria, A. R. 7 "Towards the Problem of the Historical Nature of Psychological Processes." INTERNA- TIONAL JOURNAL OF PSYCHOLOGY :-7. Perry, Milman 30 "Studies in the Epic Technique of Oral Verse-Making. I: Homer and Homeric Style." HARVARD STUDIES IN CLASSICAL PHILOLOGY :80. Reichenbach, Hans 7 ELEMENTS OF SYMBOLIC LOGIC. New York: The Free Press. Scribner, S. Scribner, Sylvia 7 "Recall of Classical Syllogisms; A Cross Cultural Investigation of Errors on Logical Problems." In REASONING, REPRESENTATION AND PROCESS. R. J. Falmange, ed., pp HillsdaTe: Lawrence Erlbaum Associates. Van Dijk, Tuen 7 "Foundations for Typologies of Texts." SEMIOTICA VI(): 7-33.