What's a rabbit?

The rabbit does not exist. If that statement suffices to summarize the content of this work, its place would not be a journal of philosophy but the office of a neuropsychiatrist. It is then necessary to clear from the outset a misunderstanding: I do not deny that there exist organisms that we call rabbits or, in particular, the species Oryctolagus cuniculus, nor that we can recognize them, study them and distinguish them from other animals. What is called into question is of another order, and its precise formulation will only be intelligible at the end of the journey. The question that opens it, however, can be stated already: what exactly do we affirm when we affirm that a rabbit exists, and to what does the use of the general term "rabbit" commit us? Before a rabbit there seems to be no philosophical problem to solve. We see an organism with certain anatomical, physiological and behavioral characteristics; we distinguish it from a rock, a tree and a dog; we use a known word to identify it. The operation is so immediate that we tend to read the familiarity of recognition as proof of something much stronger: that reality is divided beforehand into the same units that our language uses to describe it. The world would contain rabbits, trees, mountains and vessels; the words would be limited to pointing out entities already delimited before all classification. Against that reading it is necessary to separate two statements that the immediacy of recognition usually melts. The first is that there are concrete organisms, with specific bodies, that move, feed, reproduce and die, and that they resemble each other. The second is that there is something common by virtue of which they are all members of the same class. The first can be admitted without assuming an ontology of essences. The second is a substantive ontological thesis, and the first is not followed. To recognize an object does not prove that the category with which we recognize it exists in the same way as the classified object. The whole problem lies in an almost imperceptible transit between one affirmation and the other. The choice of the rabbit does not correspond, moreover, to an exhibitional whim or to the zoological formation of the writer. Since W. V. O. Quine made him the protagonist of his experiment on radical translation, this animal occupies a singular place in analytical philosophy. In the scene imagined by Quine, a linguist tries to understand a completely unknown language; a rabbit passes and the native speaker pronounces gavagai. Immediate translation seems obvious, "rabbit." But observable evidence does not allow us to decide whether the expression refers to the entire organism, to a part not separated from it, to a temporary phase of its existence, or to the event that both have just witnessed. The example shows that observation does not uniquely determine the reference: incompatible translation hypothesis can be equally compatible with the same linguistic behavior. The linguist can record when the expression appears, observe reactions, ask, compare answers; no accumulation of evidence sets in an absolute way that gavagai means the same as our word "rabbit." Here is a trait of the example that usually goes unnoticed. While we discuss what can mean gavagai, we accept without reservation that we know what "rabbit" means. Uncertainty rests on the unknown expression, not on the category with which the translator organizes what he observes: we ask if the native refers to the complete animal, to one of its parts or to a temporary phase, and we assume that what has just crossed the field is a rabbit. Even the most exotic alternatives presuppose a previous segmentation, because a part of rabbit remains part of what we already identified as rabbit, and a temporary phase remains phase of that same organism. The list of rival translations is not neutral, it is written in the vocabulary whose legitimacy should be tested. Quine did not ignore the problem. He argued, on the contrary, that the inscrutability of reference is not an anomaly of the exotic languages but begins at home, and much of his work is dedicated to examining the ontological commitments we assume in describing the world. What I propose is to take that observation seriously at a precise point, moving uncertainty from gavagai to the category of translation. It's not just about knowing what gavagai means, but what "rabbit" means and what we commit ourselves to using it: the term can be understood as a sign that allows us to classify organisms or as the expression of a real unity that exists beyond each individual. I take the argument, yes, in a restricted sense: indeterminacy affects our cuts, not the existence of the cutbacks. The suspicion also admits an empirical confirmation that comes from the least expected flank, because the category that the linguist took for granted also does not have clear limits. "Rabbit" does not name a taxon covers the European rabbit (Oryctolagus cuniculus), the American rabbits of the genus Sylvilagus and several other genera, while excluding the hares (Lepus), which however share with them the Leporidae family. Rabbits thus understood do not constitute a natural group in a phylogenetic sense. The ordinary term therefore brings together a biologically heterogeneous set, and if language were to cut reality by its natural joints it would be difficult to explain why it cuts right there. The doubt that the example directed towards gavagai thus reaches, and with better reasons, the word of translation. Before moving forward, it was necessary to grant what should be granted, because denying the essence did not oblige to deny objectivity, since the category was not self-evident did not render it arbitrary. The organisms we classify as rabbits share anatomical, physiological, genetic and evolutionary traits that can be studied regardless of the language used to describe them, and it would be absurd to argue that a rock can become rabbit by terminological decision: the properties of both material systems are objectively different and no reform of the vocabulary would alter that difference. Zoology does not gather organisms under the same category because it has decided to assign them a common word, but because it recognizes among them body structures, physiological mechanisms, phylogenetic relationships and shared genetic patterns. Scientific classifications are not constructed on an undifferentiated matter, the world presents regularities that restrict possible descriptions. We could form a category that would gather rabbits, stones and odd numbers, and its mere formulation would not make it comparable to a zoological classification, because it would not cut off any set of causal, structural or evolutionary regularities. The categories can be conventional in their formulation and be, at the same time, objectively restricted by the properties of what they describe. This concession, however, does not solve the ontological problem, and here is the core of the argument. That several organisms share real properties does not imply that the unity with which we gather those properties exists in nature in the same way that classified individuals exist. It is one thing to affirm that certain organisms have similar structures and relationships; another, very different, is to maintain that this similarity refers to a form, a nature or a common essence present in all of them. The confusion occurs when the objective character of the similarities is transferred to the category that organizes them; if several organisms are recognized as rabbits, it seems reasonable to assume that there must be something common that makes them such, a unit that does not identify with any concrete specimen but that would be realized in some way in all. The term then ceases to function as a qualifying instrument and acquires the appearance of an entity. There would no longer be only similar organisms, but also that by virtue of which they are all rabbits. It is here that the conejity is implied. It would not be any of the individual rabbits, since each one can be born and die without the category disappearing; nor the sum of the existing organisms, because we continue to talk about past rabbits, possible or still unborn. Its function would be to express what remains identical through individual differences and to explain why a plurality of organisms belongs to the same class. It is, in essence, the classic answer to the one problem over the many where there are many who receive the same preached, there must be one that melts it. The argument that leads to that conclusion can be reconstructed with some precision. Let’s write ‘Cx’ for ‘x is a rabbit’ and ‘Pxu’ for ‘x participates in u’, reserving the variables u, v for alleged non-individual entities. The premises are then as follows: (1) The first is predicative: (2a) -x -y ((Cx ~ Cy) → -F (Fx ~ Fy)) The second is reifying: (2b) -u -x (Cx ~ Pxu) Note that the order of the quantifiers does all the work. In (2a) the second order quantifier is within the reach of the universes and depends on the pair of individuals considered; in (2b) the existential precedes them and affirms a single entity for all cases. The essence appears, precisely, when that order is unduly commuted. (2a) is true and even trivial, because any pair of rabbits satisfies it: it is enough to take F = C. Its second-order quantifier might, however, seem an undercover ontological compromise, and it is then appropriate to specify how it should be read. In the preaching reading that I adopt here, "F" does not affirm that there is an entity F of which x e and participate, but that there is a true predicate of both, the range of the variable are predicative positions, not a domain of objects that is added to that of individuals. Anyone who argues that no second-order quantification is innocent can reformulate the premise in the first order by means of a likeness predicate, -x -y (Cx -cy) → Sxy), where "Sxy" abbreviates that x e and share structures, mechanisms and relationships of offspring; the diagnosis does not change, even if notational symmetry is lost with (2b). In either of the two versions, from (1) and (2a) it is not derived (3), the only thing that is obtained is that there are true predicates of both organisms, not that there is a subsequent entity of which both participate. (2b) it does allow derivation, and for an unfavorable reason, is practically the conclusion that had to be demonstrated, with the addition that introduces a new predicate, "participating from", whose extension and verification conditions remain unspecified. The argument is, therefore, either valid at the cost of a premise that asks in principle what it should prove, or impeccably in its premises but invalid; there is no third possibility. And the difference of commitment is already noticed in the singular case: «this organism is a rabbit» is formalized as Cb, while «there is the rabbitness, which makes this organism what it is» demands The first formula classifies an individual according to certain properties and relationships; the second adds an entity to the inventory of the world. Asymmetry is best seen if we look at the semantic structure of preaching. To say that an organism is a rabbit does not require that "rabbit" be the name of a thing, since it can function as a predicate that we rightly attribute to certain individuals. Similarly, to say that an object is red does not require the postulate of an autonomous entity called redness that enters the object and makes it red. We can recognize properties and establish similarities without converting each general term into the name of a separate reality and if we accept the quinean criterion according to which we commit ourselves to what we quantify, the difference becomes clear: "there are rabbits" quantifies about organisms, while "there is conejity" quantifies about a class or, worse still, about a form. None of this denies that there are real properties. The colour of an object depends on its material structure and its interaction with light; the characteristics of an organism depend on its composition, its organization, its development and its history. It is not a question of reducing everything to words, but rather of preventing the possibility of formulating a general expression from becoming, on its own, proof that there is an ontological unity that corresponds to it. The displacement is not uncommon in the history of philosophy and has been accurately diagnosed. Rudolf Carnap, in his critique of metaphysics, warned how easily the grammatical structure produces the appearance of ontological contents, and illustrated it with the Heideggerian formula according to which nothing swims. Nothingness, which fulfills a negative function within language, happens to grammatically occupy the place of a subject and, already substantive, seems able to execute an action; for Carnap we are not faced with the discovery of a deeper dimension than the real, but with an effect of the form of the expression. Moved to a less solemn terrain, the same procedure authorizes to say that the vasea vessel or that the rabbit bunny bunnies. Both expressions are grammatically flawless, and their correction does not attest to the existence of vasear or bundling as additional components of the world. Syntax allows for producing nouns, attributing preached and building deep-looking prayers; it does not guarantee that each element of prayer has an existing correlate. Where there is a nominal expression we suppose one thing; where there is a predicate, a separate property; where several individuals receive the same name, a common nature that holds them together. Rabbit is born of that mechanism, and its genesis can be described in three steps. First we observe that certain organisms share objective traits; then we gather those similarities into a category; finally we treat the unity produced by classification as if it were something different from the organisms, their properties, and the relationships they maintain with each other. The concept that served to describe a regularity ends up becoming an entity, and what belonged to the order of representation is presented as a component of reality. It could be replicated that rabbitry, even though it has arrived in this way, has an explanatory function: rabbits are rabbits because they participate in it. The supposed explanation, however, repeats in a substantive way what it had to explain. Organisms would be rabbits because they participate in bunnies, and we would know that they participate in bunnies because they present the characteristics of rabbits; the argument returns to the starting point without offering any independent criteria that establish the existence of that common entity. Circularity is hidden because the essence is presented as ontologically prior to individuals, although epistemologically we only get to it through those same individuals: we recognize similarities, abstract a common unity and then use abstracted unity to explain the similarities that allowed us to build it. The concept appears as the foundation of what originally made it possible. Nor does the reply that the essence is not a separate thing but a present form in each individual, because the defender of that thesis remains before a dilemma. If the form adds nothing to observable traits and real relationships, it becomes redundant, and Ockham's knife eliminates it. If you add something, it is appropriate to say what it consists of, what difference it introduces with respect to the simple finding that these organisms share properties and by what procedures we could know it, the second branch has never been successfully transited. There is a strong objection, and it is the one that holds much of the appeal of essentialism: without essences, the classification would collapse into pure convention. The objection confuses two problems, on the one hand a classification can be based on real properties without the conceptual unit constructed from them existing as an additional entity; the organisms can share structures, mechanisms and relations of descent without those similarities having to be unified in a form situated above them. The objectivity of preaching depends on the characteristics of individuals and the criteria by which we classify them, not on the existence of a universal. Modern biology, on the other hand, worked on this problem on its own and reached a related conclusion: since the systematic abandoned the essentialistic definitions of species, a taxon ceased to be understood as a group of individuals who share a nucleus of necessary and sufficient properties, among other reasons because there is no trait that all rabbits possess and that only they possess. What there are are populations linked by relations of descent and descent, with internal variation, diffuse limits and intermediate states, therefore the unity of a species is historical and causal, not essential. Mario Bunge’s systemic materialism allows all this to be formulated without falling into idealism or extreme nominalism. For Bunge, the world is composed of things and material systems that possess properties, maintain relationships and go through processes, while classes are conceptual constructions. A concrete rabbit is, from this perspective, a biological system whose organization emerges from the relationship between material components and historical processes; its properties are not words or fictions, but they really belong to the organism, although the concept with which we classify it is not an additional part of its body or a metaphysical form that goes through it. The category represents common properties and does not produce or exist above them. The position is maintained, strictly, on a distinction of three terms that other philosophies tend to collapse. There are (i) material systems, that is, concrete organisms; there are (ii) their objective properties and relationships, such as their anatomy, their physiology, their genetics and their evolutionary kinship; and there is (iii) the conceptual category by which we select some of those properties and gather the organisms under the term "rabbit". The first two terms belong to the studied reality; the third, to the knowledge of that reality. Almost all debates on universals arise from the sliding of the third into the first two. Two thesis follow from that distinction. The first is semantic: a general predicate does not have to function as the name of an entity, so that "this organism is a rabbit" does not amount to "this organism participates in an entity called a bunny", but can assert, more modestly, that the organism meets certain objective criteria of classification. It is therefore appropriate to separate three things that usually travel together: the objective meaning of a category, its true application to certain individuals and the supposed ontological existence of a corresponding universal. I accept the first two and deny that involve the third. The second thesis is ontological and consists of a principle of austerity: an entity must not be incorporated into the world when the material systems, their properties and their relationships already reach to explain what we observe. Rabbit does not add anything other than the common properties that organisms already possess and the concept with which we gather them; if it does not introduce any new components, it is redundant, and if it introduces something different, it corresponds to say what is and how we know it exists. There is, I understand, the strongest criticism of essentialism, and it is appropriate to formulate it unambiguously: it is not to deny the objectivity of the classification, but to deny that such objectivity requires an essence. It is worth delimiting the position also by contrast, because each of its neighbors usually presents itself as the inevitable price of denying essences. It is not a radical nominalism, since among rabbits there is much more than the coincidence of receiving the same name. Nor is it an ontological constructivism, because language does not create organisms or properties. It is not a relativism, since classifications are not all the same and reality itself provides the criteria for evaluating which best represents its regularities. It is not a realism of the universals, since in addition to the concrete rabbits there is no conejity. It is not an Aristotelian essentialism, because belonging to the category does not consist in updating a substantial form that determines what each organism truly is. Nor is it a radical quineism because the argument of the gavagai serves to show that language does not automatically segment reality, not to conclude that ontology is entirely relative to a linguistic scheme, so that the indetermination of the reference calls into question our cuts without making the world docile that we cut. A consequence of the position deserves to be underlined, because it constitutes at the same time a test in its favor. If the categories reproduced already determined ontological divisions, it would be difficult to explain why science reviews its classifications, separates groups that it considered united or brings together previously distinguished organisms. That a taxonomy can be corrected shows that it is not the immediate copy of an obvious structure, but a conceptual construction subject to empirical control and its constructed character does not make it fictitious: a map is also a constructed representation and can, however, adequately correspond to a real territory. Objectivity does not require identity between the representation and the represented, but a well-founded relationship that allows to describe, explain or anticipate properties. The category "rabbit" can represent real organisms without existing as an entity within each of them, and in that sense the disjunctive between nominalism and essentialism loses its apparent strength: it can represent real similarities, apply with truth and correct itself before the evidence without there being any form to support it. The world does not depend on our classifications, although our classifications depend in part on how the world is organized. If not, ontology would grow at the same pace as vocabulary, each noun would require one thing; each property, an essence; each relationship, an additional entity that would ensure the link between its terms. The world would be populated not only by organisms, objects and processes, but also by animals, bunnies, vassitudes and how much abstraction grammar consented, and we would have confused the richness of language with the abundance of what is. Let us then return to the initial thesis, already with its established scope. To say that the rabbit does not exist does not deny the existence of the concrete organisms nor maintains that the names are arbitrary: rabbits exist as material systems, have objective properties and maintain relationships that justify their classification. What is denied is that, in addition to those organisms, their properties and their relationships, a common entity called a conejity must be admitted. The rabbit understood as a category belongs to the knowledge with which we organize reality, and it can be an objective concept, scientifically fruitful and semantically significant without any of those virtues reaching to make it an essence; objectivity should not be confused with autonomous existence, just as the possibility of formulating a true preaching does not show that its predicate designates a thing. The question "What is a rabbit?" admits, therefore, two answers that should not be mixed. Since biology, a rabbit is an organism with certain anatomical, physiological, genetic and evolutionary characteristics, integrated into a lineage. From ontology, "rabbit" is an objectively founded conceptual category with which we gather organisms that share real properties. Confounding both planes transforms a classification into a nature and a description into an entity. Reality does not need to contain bunnies to contain rabbits: it is enough to contain material organisms whose similarities can be recognized, studied and represented by concepts. Adding a common essence does not make these organisms more real or more objective to classification, as it only duplicates in ontology the unity that we had previously built in language. Rabbits exist. Rabbits do not.