“Enjoyment and Conviviality Book” (984) by Abu Hayyan al-Tawhidi and Its Famous Dispute about Language and Logic

Full Text

The dialogue took place in 932, when Abu al-Fath b. al-Furat was a vizier¹. Many notable scholars were present. It was reported “in full” (?) by Tawhidi at the request of Ibn Sa’dan². In the following, only those parts analyzed which have a direct connection with one of the participants, the famous Sibawaihi’s “Kitab” exegete, Abu Sa’id al-Sirafi’s³ views on the Arabic language and classical logic, traced during its discussion with Yunus b. Matta, logician, Farabi’s tutor of logic. They are put and commented on following an order, defined by Sirafi’s questions to Matta and dealt with in the same consequence to clarify the famous grammarian’s theoretical views on Greek logic and, finally, his rejection of it.

1. In the beginning of the dialog, Ibn Ya’ish⁴ (as reported by Tawhidi) complains about the dire conditions for the zetetic philosophers of that time [2. P. 117]. Al-Tawhidi did not hesitate to take a defensive stance against his opponents… He says wisdom can rarely be gained through the ordinary human mind [2. P. 118]. He explains: “Whoever dares to do so will make profit out of it” [2. P. 119]. These notes were made to illustrate whether or not the one destined to be this dialog’s sole transmitter was biased. He really could be. Probably that is because it was not transmitted directly, but primarily as heard and explained by ‘Ali al-Rummani⁵, though Tawhidi heard “something” from al-Sirafi as well. Moreover, Y. al-Hamawi (in his famous “Mu’jam al-’Udaba’”) also accuses Tawhidi of “habitually romanticizing” the dialogue [3. P. 79]. It was reported only by Matta’s opponent [3. P. 86]. On the other hand, some things could be improved about Matta himself. D. Margoliouth points out the low quality of his translation, not because of imperfect knowledge of Arabic but rather due to lousy knowledge of philosophy itself [3. P. 86-87]. Nowadays, Arab scholars also see his translation of Aristotle’s Poetics as a mere “verbatim” literary work, “lacking the spirit” of the original [4]. However, it should be noted here that the Greek wisdom was at the beginning of its way there since it was introduced to study about a hundred years before by Ma’mun (813–833) [3. P. 87–88].
2. Later in the dialog, Tawhidi remembers Matta as an example of those who made hay out of it. As proof that the latter was never such a rare competent person, al-Tawhidi also cites his famous dialog with Sirafi, which will make up most of the eighth “Night”.
3. It was Aristotle’s epistemology that was raised for discussion in order to attack it with the further aim of supporting “the advocates of faith” [2. P. 120]. Generally speaking, the dialogue was destructive and Sirafi got to speak more eloquently, having the whole audience to support him, refuting everything said by his only rival. D. Margoliouth adds: “There are passages in his speech which imply that Abu Bishr (Matta – M.P) said, at any rate, rather more than he is reported to be said” [3. P. 86]. The invectives towards him are another proof of the significance of his position. Among them were “disrespect of language” [2. P. 127], “arrogance,” “elite behavior,” that Matta was a “forceless magician,” and “crafty execratory,” who should be “piped down.”
4. In the excellent tradition of agreeing on definitions, the two sides continue the dialogue by clarifying the subject of logic. Matta declares that it is like the weightier, ‘ala (which “scales” the truth from the falsehood). Sirafi says that what is weighed (the “content”) would be more valuable. By “content,” he means mostly semantics⁶, though he was supposed to be ‘defensive’ only towards the grammar only [2. P. 121]. Soon he turns to the grammar itself, claiming that ancient Greek as a philosophical language will not do for non-Greeks. Here, he resonates with the ontology of Pythagoras (where “4+4=8” is used in all languages) [2. P. 122]. Thus, Sirafi tends to strengthen the link between language and thought. This is true (according to analytic philosophy), but the latter would not have appeared without Pythagoras, Socrates, and their tradition.
5. Later, during his eloquent speech, he (Sirafi) reasonably tends towards the folk-oriented view of all “sciences” and skills which, according to him, are “equally scattered through all nations” [2. P. 124]. Sirafi even challenges Matta to master his language, explaining that it does not fit as a commentary (or a meta-level) to the Greek language or to their thoughts [2. P. 124]. Finally, Sirafi admits: “Grammar is logic, derived from the Arabic language, logic is grammar, understood within [the Arabic] language” (النحو منطق، لكنه مسلوخ من العربية، والمنطق نحو، لكنه مفهوم باللغة) [2. P. 125]. The reason, he says, is in the different value of words (as of F. de Saussure’s theory – M.P.)^[7] [2. P. 126]. This difference, unfortunately, became apparent at that time, despite the fact that Arabic grammar originally grew out of the “cross-linguistic phenomenon which never respected the boundaries we like to drew between Greek, Syriac, Hebrew and Arabic worlds,” while “being driven by its internal genius.” The author of these words, D. King, goes on: “The essentialist view of Grammar adopted by many Greek thinkers led to the working assumption that logic and grammar were virtually the same discipline, and that Syrians shared this view of things and transmitted it to Arab scholasticism” [5. P. 101].
6. Later in his speech, despite the above distinction, Sirafi could not but share many philosophical views (by also quoting from Plato’s mentalism, in particular) [2. P. 122]. That is why he hesitates. According to him, Aristotle is the only thinker who founded science and not the only one (but the Greek nation). Aristotle solved the central questions and did not. The world did not change after he applied logic. (Is that true? From K. Jaspers’ point of view?) Everything should be natural, he suggests, although he defends cultural history anyway.
7. Later, when talking about servative particles (wa-, fi, bi), Sirafi points out that wa- as a meaningful unit within the roots ("أصلية" في الاسم، كقولك: واصل واقد وافد) [2. P. 128] is still in a process of the root morphologization (like many other examples he gave, where wa- is a morpheme). Though now the prevailing meaning of these particles (in a dictionary) is that after their grammaticalization (i.e., functional meaning only, “outside” the root). Sirafi, thus, goes “too deep in the past”, when syllables were equal in expressing both functional and lexical meanings (together).
8. A page later, Sirafi challenges Matta on the level of the sentence. He poses the question as follows: which is correct? Zaid is the best of the brothers, or Zaid is the best of his brothers? In Baghdad in 932, logicians already used simple categorical and conditional-categorical syllogisms. But Rummani and Tawhidi claim that Matta could not cope with them...

But could it be? The solution, yes, is longer than the one Sirafi gave (several sentences in natural language) [2. P. 130], but the logical analysis is more illustrative. Let us see how logic can give us a more valuable explanation.

The problem is that the word ‘akh (brother) is polysemantic. The first statement “Zaid is the best of the brothers” can mean any (large) group of people, not siblings only. That is why language tries to distribute this concept. For this reason, everyone who speaks Arabic (as their first language) would prefer to form status constructus (idafa) here instead of just a definite plural. This was proved during the “linguistic experiment” in the Modern Fusha lesson with the Arab students. So, natural language does not give any other way to distribute the meaning and forces us to put it incorrectly. However, with the help of logic we can show how the volume of the word (pl. ‘ikhwa) could be taken as distributed. That is why logic is the only way to show the essence of the problem here. In fact, Sirafi does the same thing: he goes to semantics, he could not do without it in his explanation. When someone enters the field of semantics, he is partly dealing with logic, because semantics is to a large extent a logical science. Sirafi also philosophizes like a logician and defines who these brothers are. He distributes the volume of the word, which is a prerogative of logic.

First, Matta could (quickly, as will be shown below) prove the truth of one of the propositions by conditional categorical syllogism. Modus tollens (I (b)) would be more valuable here. Here is how he could do it.

Table 1
Logic solution by conditional categorical syllogism 

Source: compiled by the author.

 Second, the solution by the simple categorical syllogism (below) is more complicated. But in it we can see exactly how the concept of brothers is taken to be distributed in a certain way. We should emphasize again that Sirafi did the same procedure. Although, his explanation couldn’t be verified by any means, but only by common sense.

However, Sirafi actually agrees with Matta in all the bold statements above and in Table 2. Now, let us see how Matta could answer this question using simple categorical syllogism (see below).

Table 2
Logic solution by simple conditional syllogism 

Source: compiled by the author.

Third, nowadays it could also be illustrated by means of predicate logic^[8].

Table 3
Predicate logic solution 

Source: compiled by the author.

The only grammatical explanation that Sirafi actually gives is about status constructs (ism (idafa – M.P.) mankoor)^[9], used with nouns. Here he tries to talk about the volume of the concept, which is again subject to logic. Language taken naturally could not give a clue that Zaid is the best of his brothers is an incorrect form. On the contrary, it is the language that states it as a valid form, since it doesn’t have any other way to distribute the brothers.

We should remember that syllogism is different in Greek and Arabic logic. In the former, the logician deals with general conclusions. In the second, a traditional scholar concludes with particular statements derived with the help of analogy [1. P. 559]. Since both logics deal with their own logos, it is reasonable to compare them. In the classical tradition it is understood as speech, so closely connected with its semantics that it tries to avoid verbalization [7. P. 35]. In the traditional view, it (منطقٌ) is seen, in opposition to the previous definition, as something uttered (منطوقٌ)¹⁰ [1. P. 564].

That is because of Sirafi’s view that root phonemes are still ungrammaticalized, and he could not agree with the application of metalanguages, which are “higher” than grammar, with its terms in Arabic [2. P. 132]. This is also one of the reasons for the traditionalists’ rejection of the first Arab classical philosopher, al-Kindi. They considered his language a profound example of tongue-tiedness. (فقد حُفظ جوابُه ... على غاية الركاكة والضعف والفساد والفسالة والسخف) [2. P. 135]. The values of the Arabic words (images) were neglected there, since they were mostly taken by him only as genders (which can be distributed or, on the contrary, “emptied” with any needed meaning). Jahiz’s perspective (which can be important here) assures that philosophy cannot be accepted on the grounds of its disapproval by language! Jahiz has an invaluable record on this epistemological issue. “The Greeks have the philosophy and art of Logic, but the author of the Logic was himself a poor speaker, not regarded as eloquent, in spite of his acquaintance with the distinction and analysis of speech, its meanings, and its properties. They regard Galen as the most logical of mankind but do not ascribe to him oratory or the sort of eloquence that goes with it” [3. P. 88].

So, the question can still be asked. Was the dialogue held by Sirafi (multifunctional sheikh of the faith, imam of the believers, etc.), memorized by Rummani and then transmitted by Tawhidi as it is now? Secondly, logic makes it possible to verify certain sayings of natural languages more precisely and in different ways (I-III above). That is what led to the development of science several centuries later.

_{¹ Abu al-Fath al-Fadl b. Ja’far Ibn al-Furat (d.938), notable statesman of the Abbasside rule, was a vizier in 932 during six months [1. P. 552].}

_{² Hussain b. Ahmad Abu ‘Abdulla Ibn Sa’dan al-’Arid (d. 985), a commander of Buids’ troops. He used to invite writers and thinkers to the court in order to investigate some influential persons’ connections with different sects. In this Book he tries to find out (in “Night 17”) about one of his subordinates, Zaid b. Rifa’a’s association with the Qarmatian sect.}

_{³ Abu Sa’id al-Sirafi (d. 978), Tawhidi’s grammar teacher.}

_{⁴ Ibn Ya’ish, the Baghdadi philosopher of the late X c., close the “Brethren of Purity”.}

_{⁵ ‘Ali b. ‘Isa al-Rummani (d. 994), a famous theologian and grammar scholar, was one of Tawhidi’s mentors.}

_{⁶ Semantics in modern linguistics are seen mostly as part of the general discipline of Semiotics. Grammar (morphology and syntax) is also a part of it, but now it is considered another branch of it. Semantics were important to us here to emphasize that semantics now comprise a big deal of logic as well.}

_{⁷ The most obvious here is that by value can be understood what traditional view proposed by it: final rhyming with the end of another sura. See an example with the word deeza (53:22) [3. P. 87]. It is a legacy of saj’ style rhyming.}

_{⁸ The general idea was given by T.V. Filatov, Doctor of philosophy, Professor, head of Department of philosophy of the Povolzhskiy State University of Telecommunications and Informatics, the author of valuable article on logic in mythological texts [6].}

_{⁹ He could, though, if he had told the following. If Zaid is the best of the brothers’ equals (زيدٌ أفضلُ الإخوةِ) so the generalization would read Zaid is the best brother (زيدٌ أفضلُ أخٍ). This sentence is grammatically correct. Here we don’t need to employ any other knowledge except grammar. Here we have the pure grammatical solution. Though the generalization for Zaid as the best of his brothers would read: Zaid is the best brother of himself (زيدٌ أفضلُ أخٍ له). The second is obviously incorrect. It is also possible that this part had been lost or did not reach us.}

_{¹⁰ The idea of different kinds of Logos, including these ones, can be find in Plato’s “Phaedrus” (~360th BC).}

About the authors

Mikhail S. Palenko

Author for correspondence.
Email: m-pal@outlook.com
ORCID iD: 0000-0002-5593-0024

CSc in Philosophy, Lecturer at the Arabic language and literature Department

References