The Syllogisms of Aristotle

In the last blog post, Aristotle presented the three types of conclusions: the logical conclusion, the dialectical conclusion and the false conclusion. Now we will have to deal with the logical conclusion, which derives a true conclusion from two true premises.

This is, of course, the most important conclusion, which ensures a safe “transport” of the truth of statements. That such a possibility exists is the good news, and you can’t overestimate it. The “bad news” would then be that truth is not gained in any way, it is only passed on. The premises must be true. The question as to how and where one can begin with true statements in practice will be of great concern to us later on.

With the logical conclusion one can distinguish now again several kinds. The individual proofs, as he also called them, differ in the nature of the premises and the conclusion, which then results from a “summation” of the premises. In the Greek “summation” means “συν-λογισμός” (syllogismos); thus, one also speaks of a syllogism, the doctrine of syllogisms is called syllogistics. 

Let us first look at an example of syllogism:

All humans are mortal.
All Greeks are humans.
Therefore: All Greeks are mortal.

So, there are three statements, two propositions from which one proceeds, and a conclusion in which the propositions are “added together”.

The form of sentences and their representation

In this example, all sentences are of the form “All A are B”.  The sentence “All A are B” can also be formulated as follows: “B applies to all A.”  This formulation suggests that B is a predicate of all A: B is attributed to all A (predicated), e.g.: To be mortal applies to all human beings.

The formulation is also the one that comes closest to the Greek text, so it is the more original form. In scholasticism however, this was rewritten then as “All A are B”. All A has the predicate B: All humans are mortal. 

The above syllogism is then in its original form:

To be mortal belongs to all human beings,
To be a human being belongs to all Greeks.
Therefore: To be mortal belongs to all Greeks.

Moreover, this formulation also suggests a method that was discovered much later to illustrate the relationship between two terms. One takes advantage of the fact that each term applies to a set of entities. “Human beings” can be Greeks but also Egyptians, Thracians or Asians or Europeans. If one represents the set of human beings by a circle or any closed curve (B), the set of Greeks as well (A), then the circle for the Greeks lies within the circle for the human beings.

Fig.1: The set of Greeks (A) is a subset of the set human beings (B). To be a human being (B) applies to all Greeks (A).

Such pictures are called Venn diagrams, after the mathematician John Venn (1834 to 1923), who introduced them following Leonard Euler (1707 to 1783). Actually, the philosopher Gottfried Wilhelm Leibniz (1646 to 1716) already used them. These Venn diagrams thus illustrate relationships between two sets in general, no matter of what kind the elements are. Within the framework of set theory, we would also write: A ⊂ B, i.e. A is a subset of B.

But this form of a sentence is not the only one at syllogisms. Aristotle obtained an overview of all possible forms of sentences and statements. This resulted in: (Prior Analytics, Book I, 1,  see
https://ebooks.adelaide.edu.au/a/aristotle/a8pra/book1.html):

A premiss then is a sentence affirming or denying one thing of another. This is either universal or particular or indefinite.
By universal I mean the statement that something belongs to all or none of something else;
by particular that it belongs to some or not to some or not to all;

by indefinite that it does or does not belong.

Thus, beside the general sentences like “All Greeks are human beings” or “To be human belongs to all Greeks” there is also the negation (no-compliance) and the particular statement. Altogether you get the following types of sentences:

B belongs to all A, (all A are B),
B doesn’t belong to any A, (no A is B),
B belongs to some A, (Some A are B),
B doesn’t belong to some A, (some A are not B).

We can illustrate the other types of statements as follows:

 Fig. 2: Left: B doesn’t belong to any A. Middle: B belongs to some A. Right: B doesn’t belong to some A.

In scholastic circles these forms are abbreviated as (A a B), (A e B), (A i B) and (A o B). The letters “a” and “i” should remind us of “affirmo”, “e” and “o” of “nego”.

The structure of syllogisms

Two such forms of statement then form the premises, one such the conclusion. Each conclusion can thus be characterized by three letters from the set {a,e,i,o} and by the position of the three letters, which stand for the concepts in the respective sentence.

One of these terms, the so-called middle term, must occur in both conditions, it can be in first or second place, or in one condition in first, in the other in second. This results in four different shapes or figures. The first figure is the following (see example above)

A – B

B – C

A – C.

Now Aristotle selects in all forms that combination of two forms of statement that necessarily lead to a conclusion. He simply sorts out the conclusions for which he finds a counterexample.

The valid conclusions can then each be represented by the form and a specific combination of the letters a,e,i,o. And in order to be able to remember such combinations better, one has integrated them into corresponding peculiar words, e.g. one remembers the combination a a a with the word “Barbara”, and knows in addition that here the 1st figure is present. This conclusion corresponds exactly to the above example.

Fig.3: The Barbara syllogism

Another important conclusion, also of the 1st figure, is called “Celarent” in this way:

Fig.4: The syllogism Celarent

Aristotle demonstrated in this way that one could systematically formulate a system of conclusions in which from the truth of the premises necessarily follows the truth of the conclusion.

With the systematics of Aristotle, one now has a complete overview of all possible conclusions. Previously however, in syllogistics, “the whole art of syllogistics had consisted in searching around with great effort of time and effort”. as Aristotle wrote in his work On Sophistical Refutations (after Schupp, I 275).

But here one already has a system of statements that is reminiscent of the formal predicate logic that will be introduced later. Like there, you can “quantify” using the predicates, i.e. you can operate with quantities such as “all”, “none” and “some” for the predicates.

Most importantly, Aristotle must select the valid conclusions “by hand” from the set of all possible combinations, simply by discarding those that he recognizes as invalid with the help of an example. This “recognition” is an intuitive one, one with “common sense”. One does not doubt the correctness of the conclusion, but for a strict science in today’s sense this kind of knowledge is not sufficient.

Even if one describes the relations of terms mentioned above with the help of set theory, and thus justifies the conclusions within the framework of set theory, the insight is a mathematical one. Thus, the conclusions would be only indirectly logically justified, because mathematics only uses, as we know today, the rules of inference that are ultimately gained in modern predicate logic. Only in this predicate logic can the conclusions be strictly justified by deriving them from tautologies. We’ll see about that in a later blog post. Only in this logic is the bottom reached on which one can incontestably win a true sentence again from true sentences.

We still would have to comment on Aristotle’s remark on the case that the conditions are only likely to be true or credible. Then the conclusion could also only probably be true. Saying more here was not possible at that time.  It was not until the beginning of the 20th century that a theory of probability was developed with which one can become more precise in this case. This will be explained in more detail in a later blog post on the topic “How to deal with insecure knowledge”.

Syllogisms in an axiomatic-deductive system

But Aristotle’s logic has not only shown us how safe knowledge can be passed on safely. He has also shown that “It is possible also to reduce all syllogisms to the universal syllogisms in the first figure.”, as he writes in the Prior Analytics, 1st book, 7th chapter (see https://ebooks.adelaide.edu.au/a/aristotle/a8pra/book1.html).

Thus, we already have a concrete axiomatic-deductive system. This logical order of the statements in a field of knowledge represented for him the ideal of a science. In the Posterior Analytics, 1st book, 3rd chapter he writes:

On the other hand, I maintain that any science must be based on proofs, but that the knowledge of the unmediated principles is not provable. And it is clear that this must be necessary. For since a knowledge of the earlier propositions from which the proof is made is necessary, but one stops once at unmediated propositions, these must necessarily be unprovable. This is my view and I maintain that there are not only sciences, but also supreme principles of them through which we learn the concepts of conclusion. (Here I prefer the translation of a german version in http://www.zeno.org/Philosophie/M/Aristoteles/Organon/Zweite+Analytiken+oder+Lehre+vom+Erkennen/1.+Buch/3.+Kapitel).

The mathematician Euclid of Alexandria logically arranged the geometric knowledge of that time in this way in the time around -300. This organization of a scientific theory as an axiomatic-deductive system is still a model for any rigorous science today. One tried in all centuries to imitate this organization of a thought building, thus “more geometrico”, after kind of the Euclidean geometry, to arrange the knowledge of his science. In “Die Idee der Wissenschaft – Ihr Schicksal in Physik, Rechtswissenschaft und Theologie” I described how successful it has been to this day to realize this idea (Honerkamp, 2017).

Who first had this idea is not clear. The mathematical proof was already known to the Pythagoreans. If there are enough statements in an area, secured by proofs, one will probably at some point consider which statements could be regarded as axioms. This poses the question: Which statements do I need as a basis in order to be able to deduce all the others from them? Or also: How do I create a logical order?

Such specifications will then always be special statements, have special properties. They can be immediately insightful, i.e. “certain by themselves”, as in theories of mathematics, but also highly abstract and far from our idea, as in theories of physics.