Whitehead’s philosophical approach is based mainly on descriptive generalization. In this respect, he considers deduction as the method of mathematics, and the borrowing of this method by philosophy an error for the latter’s method of descriptive generalization (Whitehead 1978, p.10).
Indeed, philosophy starts with concrete situations and experiences which it attempts to explain. It sees order everywhere and endeavors to comprehend the origin of this order. Therefore, philosophy assumes principles that act everywhere, in the same way, to produce order in the world and, thus, to express themselves within this world.
Generalization is the method through which we extend the results of a set of particular observations to the whole class of the members related to that observation.
For example, we see that lions eat flesh and thus that they need to kill other animals in order to satisfy their hunger. From this observation, we generalize that all lions eat meat, and we assume that in the future, whenever we see a lion, it will be an animal that eats flesh. The same is true concerning other aspects and contents of reality.
Because we saw in the past that people’s actions are led by values, we assume that all humans living in communities behave according to values that they hold. Values can differ from one community to another. However, behaviors, in general, are always motivated by values.
Whitehead does not speak only about generalization but about descriptive generalization. This means philosophy describes and generalizes. Unlike sciences that ignore description, grounding their methods, especially on abstraction, abstract principles, and concepts, philosophy uses abstraction to a lesser extent, and from a particular point of view. It does not entirely ignore most of the concrete qualities of the real objects, as sciences do, especially mathematical sciences.
Aristotle said in antiquity, that science is knowledge of the universal; as a consequence, each kind of theoretical knowledge involves abstraction, i.e., the attempt to discover and describe a general structure actively present in all material and concrete realities.
Whereas mathematical sciences retain only numerical structures – which are, according to Whitehead, the most abstract structures – philosophy tries to describe perennial aspects of reality that are not as abstract as numbers.
For example, when Aristotle defined the human being as a rational animal, he maintained animality and rationality as defining features. To understand these features, they have to be described starting from the immediate objects that we meet within our experience.
This does not happen with numbers. The latter are abstract entities that are understandable through definitions. In order to conceive of animality, we must observe a lot of animals and extract their common features. The described features are general.
However, they are far less general than numbers. If we say that in a room there are 70 people, the number 70 can be applied to those people but also to many other things: we could speak about 70 leaves under a tree, 70 cars on a highway or in a car-park, 70 trains and so forth. In all these cases, we speak about the same number, but, except for quantity, no common feature connects all these things.
We need not to observe the particular features of all the 70 things to encompass them in the same class. Of course, when we do this, we need specifying features. However, in a class based only on numbers, those specified features are ignored: they are only conditions making our numbering possible. This is why numbers are made of units that ignore all other features of the items brought together in that class except that they are units in themselves.
We cannot do this with philosophical generalizations that also involve descriptions of general features. These different features can no longer be ignored if we are to have a real comprehension of what we call generalized features.
On the other hand, philosophy cannot base its method exclusively on deduction (Whitehead considers deduction only an auxiliary method, p. 10), because it tries to understand the world as a place where humans live, and satisfy their most fundamental interests. Therefore, it must observe and research this world as much as possible and create a generalization that is abstract enough to be applied to all components of reality.)
Each philosopher discovers new general features of the world or distinct aspects of it, increasing thus its overall intelligibility. This is why, we might add, the philosopher’s work is essential. The generalities of philosophies are not like the generalities of sciences. They are always related to the most fundamental (spiritual) interests or questions of the human being that do not change through the ages.
Because these generalizations consider the world as a whole (not only some parts of it, as sciences do), philosophical generalizations cannot be replaced by any fundamental scientific scheme. From this point of view, the latter is always much more limited than philosophical generalization.
In this respect, we might say that, however general the explanatory schemes of physics are, they cannot operate in domains like literature or (human) history. We cannot explain literature as a result of the motions of subatomic particles.
Those who claim that someday we will find the immediate connection between a literary work and the motion of those particles only speculate, expressing thus a belief; they cannot offer a rational argument for their belief.
Whitehead considers that the difficulty which has always confronted philosophy is the need to elaborate a generalization while explaining the huge diversity of the world, starting from a language developed in relationship to those diverse and material things. The words of our languages denote immediate things; for each class of things, we have a particular word.
The increase of knowledge always means the discovery of common features between such classes, thus moving away from immediate experience and climbing more and more into the realm of abstraction.
The endeavor to discover features specific to all types of existence involves the highest effort of abstraction associated with the need for language creation: now we deal with entities for which we have no common words since they do not relate only to our everyday experience.
Alfred North Whitehead, Process and Reality. An Essay in Cosmology, The Free Press, New York, 1978