On the contrary, for illustrative purposes we are always limited to highly restricted fields, such as furniture to sit down on or parental relationships. This does not exclude the possibility of our constructing one day (in theory) a global system of content (nor does it exclude the existence of such a system in the Divine Mind). All it tells us is that (precisely because, as Kant used to say, empirical concepts can never exhaust all their determinations) we can proceed only by temporary settlements and successive corrections.
Even the observation sentence This is an egg depends on cultural conventions. But, though egg and mammal are both concepts that spring from a cultural segmentation of the content, and though the very concept of mammal also takes account of experiential data, there is a difference in the proximity of the construction of the concept and perceptual experience (and it is on this that the difference between NC and MC is based).
When we say that, in order to decide whether an animal is a MAMMAL or not, we have to fall back on a system of cultural conventions (or, as we have seen, reconstruct one), while, in order to decide if something is an egg, we intuitively put our faith in perception and an elementary knowledge of the language being used, we are saying something that goes beyond intuitive obviousness. Of course if someone has not been trained to apply the word egg to a certain CT (which already considers the form, the presence of yolk and albumen, the presupposition that if this object is sat on for the right amount of time, then a living creature might be hatched from it), there will be no agreement on the recognition of an egg.
Therefore perceptual consensus too always springs from a prior cultural agreement, no matter how vague or folk it might be.15 And this confirms what I was trying to say shortly before, that in the process of understanding, the structural moment and the interpretative moment alternate and complement each other step by step. Nevertheless it cannot be denied that, in defining an egg as such, the testimony of the senses prevails, while in order to define a mammal as such, what prevails is a knowledge of classifications and our agreement on a given taxonomic system.
When we go on to refine perceptual judgments, and there is a clash about whether something is milk or mucus, it is necessary to treat the perceptual experience in cultural terms too and to decide which criteria and chemical classifications allow us to distinguish milk from mucus. But yet again we have evidence of an oscillation and a constant complementarity of our two ways of understanding the world. At a given moment even the common CT that would have allowed Saint-Hilaire to recognize something as milk had to give way to an MC already imbued with structured oppositions, on the basis of which it was inevitable that Meckel and Owen would emerge victorious.
4.6.3 Where does the Amorphous Continuum Lie?
All this brings us back to the opposition between the systematic, or holistic, pressure of a system of propositions and the possibility of observation sentences dependent on perceptual experience.
The postulating of a perceptual semiotics ought to give rise again to the rift between those who maintain that we give form to an amorphous continuum, and that this form is a cultural construct, and those who maintain, on the other hand, that what we know about the environment is determined by characteristics of the environment itself, from which we take the salient information it offers us sponte propria.
It seems obvious that even an observation sentence such as It’s raining cannot be understood, and judged true or false, if not within a system of linguistic conventions on the basis of which we distinguish the meaning of rain from that of mist and dew, and that therefore the concept of “rain” depends not only on some lexical conventions but also on a coherent system of propositions regarding atmospheric phenomena. To use a formula that Putnam attributes to West Churchman, who attributed it to A. E. Singer, Jr., who in his turn meant it to be an efficacious condensation of James’s thought (Putnam 1992: 20), “knowledge of facts presupposes knowledge of theories, knowledge of theories presupposes knowledge of facts.” However, the meaning of rain does not depend on the chemical notion of water; otherwise the unlearned could not assert that it is raining, and each one of us would assert this falsely in the case of “acid rain,” in which God only knows what is falling from the sky. In the same way, in order to observe that it is sunny, or that there is a full moon, it is certainly necessary to share a sort of segmentation, albeit an ingenuous one, of the astronomic continuum, but it is not indispensable to know the astrophysical distinction between star and planet.
An ingenuous segmentation of the continuum can also survive within a system of interconnected notions that actually denies it: this is why we have no problem in asserting that the sun rises, when in the light of the system of notions upon which our knowledge is based, we ought to know that the sun does not move in that way.
Let us try to imagine an imaginary debate between Galileo, one of his Ptolemaic adversaries, and someone who prefers to keep one foot in both camps, such as Tycho Brahe, Kepler, or Newton. I don’t think we need an enormous amount of imagination to assume that all the participants will agree on the fact that at a given moment they can see the sun or the moon in the sky, that both bodies seem circular in shape and not square, and that they illuminate over Arcetri something that everyone can recognize as trees. Nonetheless, within diverse systems of propositions, movement, distances, functions of the sun and the moon, notions such as mass, epicycle, deferent, gravity, or gravitation not only assume a different value but also can be acknowledged or refuted. However, even if each debater has a different conceptual frame of reference, all of them perceive some objects and phenomena in the same way.
For the heliocentrists the sun’s movement is apparent, whereas for their opponents it is real. But this difference is relevant with respect to a coherent system of propositions regarding the universe, not with respect to the observation sentence upon which both parties have agreed.
It is one thing to ask whether everybody sees an eclipse of the moon and another to talk about the movement of the heavenly bodies that produce the perception of the eclipse. The first problem concerns the way in which we form a perceptual judgment (which despite its being dependent on the structure of our cognitive apparatus must nonetheless account for the manifold of sensation), while the second concerns a system of propositions (for Kant, a system of judgments based on experience) that is certainly influenced by internal structural relations. When we talk of holism, we mean the solidarity of a system of propositions; when instead we talk of perception, even though we can presume that it is influenced by a system of propositions that create a series of expectations, we are talking about observation sentences that must in some way take account of what the environment is proposing to us in an immediate sense.
I am well aware that advocating the existence of observation sentences independent of a general system of propositions was said by Davidson to be the third dogma of empiricism; but we cannot ignore the evidence that it is easier to negotiate (in a very short time) our assent to the sentence Watch out, because there is a step than to the sentence that expresses the second law of thermodynamics. The difference is that in the first case I immediately run a check on perceptual bases (the concept of step is an “empirical concept”). Thus in the story I told in 3.5.1, Gabriel and Belphagor might have had very different notions regarding virtue, but both were able to tell the sexual difference between Joseph and Mary.
Therefore, even if we admit that every cultural system and every linguistic system upon which it rests segments the continuum of experience in its own way (Davidson would talk of a “conceptual schema”), this does not alter the fact that the continuum organized by systems of propositions already offers itself according to a grain that provides directives for intersubjectively homogeneous perception, even between subjects that refer to different systems of propositions. The segmentation of the continuum brought into being by a system of propositions and categories in some way takes into account the fact that that continuum is no longer entirely amorphous; in other words, while it is propositicmally amorphous, it is not entirely perceptually chaotic, because within it, objects interpreted and constituted as such on a perceptual level have already been carved out: as if the continuum in which a system of propositions carves out its own configurations has already been tilled by
