A Deeper Notion of Scientific Object

Our discourse on scientific objectivity is still far from being concluded with the (important) acquisition consisting in recovering the legitimacy of scientific truth thanks to the referential purport of operations.

In order to overcome the radical empiricist prejudice we must first reject the tempting idea that scientific objects are things, despite the fact that such an idea seems to be the best support for scientific realism. Indeed, in our preceding con­siderations we have pointed out: (i) that a thing can become the object of a given science as far as it is considered from the specific “point of view” of that science which determines a particular “clipping” within this thing; (ii) that one single thing can be the object of an indefinite number of scientific investigations. Therefore, equating a scientific object with a thing would amount to ignoring the distinction between thing and object that has been central to the whole of our analysis.

But now it is precisely such conceptual structure that—after having been suit­ably refined and made totally explicit—becomes the effective object of the special science considered. Therefore, this is an abstract object that, in order to be investigated, must also be denoted by an appropriate structured set of predicates constituting the specific technical language of the science in question. Mass-point, uniform motion, frictionless motion, rigid body, perfect gas, adiabatic process, isolated system, insulating and conducting material are familiar examples of such abstract objects studied in physics, while chemical reaction, chemical equilibrium, metabolism, reproduction, homeostasis, perfect competition, marginal utility, per capita income, marginality, demographic transition are examples of abstract con­cepts elaborated respectively in chemistry, biology, economics, sociology. The theoretical dimension of a science consists indeed in the construction, articulation and development of such abstract objects that have the statute of intellectual enti­ties, endowed with a quite precise sense and a logical structure. For this reason they are different from nothing, they are the content of thinking, they have a kind of reality that we could call mental or “noematic” (using a Husserlian terminology) and they are the scientific objects in a proper sense, because they are what a science directly investigates.

Empirical sciences, however, aim at investigating their own domain of objects (as we have noted) and this is not meant to be just a realm of abstract entities having a mental reality. How this can happen can be clarified by distinguishing encoding and exemplifying. An abstract object “encodes” a certain number of concepts that characterize it in an explicit and exact way, but it can be “exemplified” by many concrete things that are endowed with those attributes that are encoded in the abstract concept, and this normally occurs only within certain limits of approxi­mation or tolerance, depending on several practical considerations. This happens because the same thing normally exemplifies many abstract objects (or concepts) and these simultaneous exemplifications inevitably entail only a partial satisfaction of any particular abstract concept. For instance, an iron bar is a good exemplifi­cation of the abstract concept of rigid body, however it is not ‘perfectly’ rigid and also presents some elasticity (something that may be very useful in certain concrete applications). We call referents the entities exemplifying an abstract object, and here we see that the objects of which we have spoken in the introductory part of our discourse when we have said that every science investigates only its own domain of objects are more properly called the referents of that science, that has its proper domain of referents. It is important to recognize that the two domains—that of the abstract objects and that of the referents—are deeply different and not just a kind of a reduplication of one other. For instance, on the one hand, no abstract concept enjoys the properties that it encodes (the concept of uniform motion is not in motion, the concept of a quadruped animal has no legs, etc.) and, on the other hand, no concrete thing encodes properties, because it is not completely characterized by any finite set of properties. The relation between encoding and exemplifying has represented a complex issue that has challenged the mind of great thinkers starting at least with Plato. Without entering such a difficult problem, however, we can propose a practical criterion for determining the referents of abstract objects: this is the use of standardized operations that, in the sciences, play the role of criteria of referentiality. For a correct understanding of this point it is useful to make a short digression regarding a fundamental issue of philosophy of language and semiotics.

6