On Interpreting Something as Food

In this paper we discuss the role that individual and collective acts of interpretation play in shaping a metaphysics of food. Our analysis moves from David Kaplan’s recent contention that food is always open to interpretation, and substantially expands its theoretical underpinnings by drawing on recent scholarship on food and social ontology. After setting up the terms of the discussion (§1), we suggest (§2) that the contention can be read subjectively or structurally, and that the latter can be given three sub-readings. We then lay out (§3) three case studies that, we submit, any viable theory of a metaphysics of food should be able to account for. We show that one structural reading—based on the idea of negotiation—swiftly accommodates for the three case studies. We thus conclude that this reading is most promising for charting a metaphysics of food.


Interpreting Food
In a discussion concerning food aid programs, it seems plausible to frame our discourse in terms of basic items such as rice, potatoes, cassava, or apples thinking of such staples as food for humans in general. At the same time, it seems equally plausible to claim that what constitutes a food item is strikingly personal and contextual. In fact, for any edible item, typically within the same society, we can find different dietary, aesthetic, and value-laden perspectives. Sometimes even the very same eater will shift perspective over time or from context to context-e.g. a person eating meat as a child due to family upbringing will turn vegetarian during adulthood. Can a metaphysical theory of food account for this wide range of apparently conflicting perspectives, without defying other intuitions about what generally speaking is regarded as food for humans?
In Food Philosophy, Kaplan contends that food is always "open to interpretation" (2019: 2) and that its identity "depends on how we conceptualize it" (12). Thus, "something only becomes food for another, who interprets it as a food" (15). At the same time, Kaplan warns, any food is partly a social and partly a natural entity (15). How can this be?
According to Kaplan, anything that is food must be a natural entity. Natural entities can be characterized in terms of natural kinds or in terms of the sort of properties they possessnatural properties-including biological, chemical, or physical properties. 2 Any typical material food is associated with a bunch of those properties, which are constitutive of its identity, in fact as Kaplan (12) claims food is equipped with "real properties that are independent of what we think or do," insofar as it holds "a mass, volume, and chemistry." For instance, wine must follow a fermentation process, eggplants and tomatoes are the berries of plants that belong to the same taxonomic category (Solanaceae), and so on.
While food is a natural entity, which properties would count as constitutive of a given food, if any, is open to eaters' interpretations, as constitutive properties vary across communities, groups, and individuals. As Kaplan puts it, "hunted and gathered foods […] are just animals and plants until we decide to make them food" (15). Now, not all interpretations of the natural properties of food turn out to be equally plausible. For instance, chalk, paper, and kibble can hardly be interpreted as food (3). Thus, Kaplan concludes, there is no fixed interpretation shared by all the eaters. On the contrary, eaters individually and collectively do bear the power of deciding what is food. Here the notion of interpretation takes central stage in a metaphysical theory of food.
Drawing on the hermeneutic tradition which traces back to Gadamer (1989), Kaplan sees an interpretation as a practice against a theoretical background with one or more practical outcomes. Thus, with food, the theoretical background of the eater specifies which properties make an entity into a food. The background comes to form via a collective training within a community (18). Such training delivers what are technically called prejudices, which are backed up by individual creativity, providing the subjective contribution to the interpretation. According to Kaplan, then, the entity in itself is not a food, not because it lacks the constitutive properties for being a food, but because no eater's background enables interpretations that embed such constitutive properties yet.

On Subjectively and Structurally Interpreting Something as Food
So far, the theoretical role of interpretations seems relatively straightforward. The difficulties arise when we start wondering what grounds the existence of interpretations. How do they come to be? Which agents and acts give rise to a given interpretation? As we see it, there may be two main theoretical readings of interpretation, a subjective one and another based on social structures. We elaborate on them in the remainder of this section.
According to the subjective reading, interpretation is an intentional act exerted by a conscious eater toward a food. The model is simple: the natural domain is full of things equipped with a vast array of properties. Whether or not those properties are food's constitutive properties is laid down by the eater who can turn their bearer into food on the basis of their perspectives. The model has the resource for explaining why there is so much diversity among different eaters and different cultures since they are rooted in different ways of conceptualizing the world. 3 In many passages, Kaplan seems to lean towards subjectivism insofar as he deploys the metaphor of dialogue-which analytically requires conscious eaters-for cashing out the character of the interpretation. Thus, elaborating on Gadamer, Kaplan states: "understanding for Gadamer is like a dialogue (...) we understand things in response to implicit questions by testing our interpretations in an open-ended dialogue with others" (17).
At the same time, the rest of Kaplan's text is also in keeping with a structuralist reading of interpretation and we suggest that this reading seems more promising. According to it, interpretation rests on the idea that an individual interprets a natural thing N as a social thing S insofar as they abide by a certain social structure according to which N is S, e.g., a piece of paper is also money (Guala and Hindriks 2015). When it comes to food, then, a corresponding formula goes like this: an individual interprets a natural thing N as a food F insofar as they abide by a certain social structure according to which N is F, e.g., a fruit of the apple tree is also a food.
What is a social structure? Providing a full picture encompassing languages, norms, habits, as well as specific paths of perception and metabolism is overly ambitious for this context. We shall rather concentrate on those aspects clarifying how structures relate foods with interpretations. At the cost of oversimplifying an articulate offer of theoretical positions, we shall divide up the field into three main families of replies. To do so, we employ two classic metaphysical questions, which trace back to Aristotle and were famously put to use by Thomas. Of any given thing, we may ask Does it exist? and What is it? According to the so-called received view, the first question pertains to ontology, which is the study of what there is, while the latter to metaphysics, which is the study of what is what there is. 4 In a structuralist framework for food items, we would then ask, of a food, in what ways its existence depends on the existence of some social structure and in what respects its identity conditions depend on some social structure. Three families of replies can be envisaged here.
Global Structuralism According to Global Structuralism, all that exists are human-dependent structures, e.g., linguistic, social, conceptual, perceptual, and so on, and all the features of these structures depend on humans' interpretations. For instance, rocks, stars, cows, bodies, exist in virtue of the existence of a certain set of structures and all of their properties, e.g., being hard, edible, red, depend on those structures too. Food is no exception and it is singled out within a structure which, in turn, either creates its components or assigns to the components of an other structure, say, the one delivered by biological science, the food's nature with all its subsequent properties, e.g., being edible, being nutritious, and so on. In a nutshell, Global Structuralism claims that: (i) natural things existentially and constitutively depend on a social structure; (ii) the interpreted thing being's food existentially and constitutively depends on a social structure. Tentatively, one may put within this family of positions Nietzsche's perspectivism (1967), Derrida's deconstructionism (1976), and Goodman's irrealism (1978).
Local Structuralism Local Structuralism is famously exemplified through the metaphor of cookie cutters: while the dough is independent from the cutter and the cook, the shape and the properties of the cookies depend on the cutter and the cook (Putnam 1988). Outside of the metaphor, the ontology is independent of human interpretations, but metaphysics is not (the properties conferred to what there is depend on human interpretations). That is, there is a large amount of non-differentiated stuff which fills the spatio-temporal framework in which we live in 5 and our structures carve out rocks, stars, cows, bodies, by singling out them and by imposing on them specific properties, either by creating those properties, e.g., social ones, or by selecting the more salient features of the non-differentiated stuff, e.g., mass, charge, molecular structure. Likewise, the fact that a thing is a food trades on a structure, while its existence depends on the extant existence of the stuff from which it has been carved out. That is, say, the stuff of which an apple is made of exists independently from a structure but its being an apple, and then a food, its being nutritious, and its being edible, depend on a structure. This family of positions is tentatively exemplified by Putnam's internal realism (1987) as well as Kant's (1999) and Husserl's (1973) transcendental philosophy.
Negotiational Structuralism Negotiational Structuralism holds that some natural things dictate some constraints to structures. When interpreting the world, we "bump" into negative spaces of interpretations, which cannot be explored-the so-called "one way" "or no entry" of being, as Varzi puts it (2011b). Thus, whether certain things exist qua food and what properties they hold in order to be a food are negotiated between those things and structures as well as between different structures. Negotiational Structuralism introduces the notion of negotiation, which replaces the notion of dependence and which relates things, food, and structures. The existence as well as the identity of food is constitutively negotiated within such structures and also between interpreters who are linked to different structures. There is no longer a subordinate relation between things/foods and structures but instead a pear-to-pear relation. As Hacking (2002) famously puts it, there is indeed an interaction between the interpretation and the interpreted thing-what he called a looping or a feedback effect-which respectively shapes both. This effect here can be rephrased in terms of a permanent negotiation between structures, independent things, and interpreters. Negotiation is, in fact, twofold, inasmuch as it recognizes multiple ways of carving foods out of reality and also the mutual relation between different structures. This position owes a debt to Umberto Eco's negative realism (1990), to the work of Smith and Varzi (2000) on fiat and bona fide boundaries 6 as well as to the modest realism of Kitcher (2001).

Testing Interpretations
We now assess Subjectivism and the three versions of Structuralism by means of three tests based on three case studies: the unaware and sidestepped eater, the absentee enzyme, and the antagonist interpreter. All of these case studies involve some epistemic misalignment, either between the interpretations of different agents or because of a mismatch between the interpretation of an agent and certain facts it addresses. We show that Subjectivism cannot suitably accommodate the tests, that Global Structuralism faces perplexing difficulties, and that Local Structuralism can hardly accommodate for the absentee enzyme test and lacks a general strategy to address the antagonist interpreter test. We conclude that Negotiational Structuralism seems the most promising avenue to ground a metaphysics of food on a structural view of interpretation.

The Unaware and the Sidestepped Eater
The first test concerns cases in which the actors involved are not (totally) aware of whether what they are eating is food. Consider a three months old infant who is fed on milk formula. The milk cannot be interpreted as food by the infant insofar as they lack the cognitive capacities of interpretation, and of course they do not even have the social and cultural relevant 6 "There is a basic distinction, in the realm of spatial boundaries, between bona fide boundaries on the one hand, and fiat boundaries on the other. The former are just the physical boundaries of old. The latter are exemplified especially by boundaries induced through human demarcation, for example, in the geographic domain." (Smith and Varzi 2000: 401). abilities for providing an interpretation whatsoever. The case of infant food, of course, generalizes to other cases of unaware eaters. 7 Alongside it we can put other cases of agents who-temporarily or permanently-lack the ability of interpreting their actions or surrounding things, such as people who are in a coma. Or, we can cite the case of people who may be charged with a defective interpretative ability, such as those who suffer from alzheimer and, in extreme cases, patients with anorexia nervosa who refuse the food but are force-fed by courtorder. We may even add to the list people who are mindlessly eating, without paying attention to what they are shoveling through their mouths. Finally, we should probably also include paternalistic cases, such as those of the patients of a hospital, whose interpretation is sidestepped by medical opinion.
Generally speaking, the argument shows that in some cases, x is fed by y and nevertheless x does not interpret y as food. Why this happens lies in many sorts of explanations about the eaters and their context, e.g., minority, lack of information, unconsciousness, recklessness, indifference. Thus, if we admit that, in order to be a food, something should be interpreted as such, should we also concede that interpretation cannot be a necessary condition for being food since in a number of cases it does not occur?
At risk of pointing out the obvious, we contend that Subjectivism lacks a straightforward strategy for accommodating our semantic practices and our ontological expectations with respect to what counts as food in all these cases. The cases of the unaware and of the sidestepped eater show that it is not infrequent that certain natural things are food regardless of the subjective interpretation of the eaters as well as all the human dependent structures that inform a global structuralist perspective.
Global Structuralism would face substantial difficulties in explaining the cases too. A global structuralist may suggest that, in these cases, the interpretation that matters is the one of the person who is in charge of making the dietary choice, e.g., the nurse in the case of the infant or the doctor in the case of the hospital patient. Yet, it is possible that not even the nurse or the doctor consider the entity as food, but rather follow the dietary advice under the prospects of some economic profit. Should a defender of Global Structuralism bring in some additional perspective, which interprets the entities in question as food? But, why should one perspective be valued more than another? Or, should they simply conclude that unaware and sidestepped eaters are not eating food, at the cost of clashing against current semantic practices and ontological expectations-not only those adopted in ordinary language, but also e.g. in pediatrics-that regard milk formula as food for and to the infant regardless of the infant's interpretation? We are not arguing that Global Structuralism may not be able to address these cases, but certainly it faces some substantial theoretical challenges.
Local and Negotiational Structuralism, instead, can punctually explain cases such as the unaware and the sidestepped eater. In fact, both can rely on the fact that there is no worrisome ontological disagreement, rather there are just different and equally plausible ways of carving out reality (Local Structuralist) or that there are diverging ways of carving out reality that call for theoretical negotiation (Negotiational Structuralism). Either way, the complexity of the case studies can be addressed without running the risk that there is no underlying common reality. After all, Local and Negotiational Structuralism are packed with a strong background constraint: ontology depends also on non-interpretative facts. The existence of something that is ingested and that triggers certain metabolic reactions is taken for granted. Whether or not that thing is food is demanded to structures and not to a subject who should posit a first ontological interpretation.

The Absentee Enzyme
This second test draws on the seemingly plausible intuition that part of what an eater regards as food is due to biological factors lying beyond the scopes of interpretation. Consider the case of the enzyme lactase. As is well-known some people lack such enzyme, which helps break down lactose during digestion. This condition may be due to different causes and may be wellknown to those having it. For those suffering from it, long term and continuous milk consumption may increase the risk of major diseases. Nonetheless, these may insist on feeding on milk, which they interpret as food despite the medical condition. Is milk food for them? Or: does interpretation alone turn a natural thing into a food, in their case?
The case of the absentee enzyme presents us with a situation where a certain biological condition is necessary-even if not, by itself, sufficient if we buy into an interpretivist framework-in order for something to be food. The case naturally generalizes to many other medical conditions, including people with coeliac disease, diabetics, patients with Crohn's disease, and so on.
As it turns out, both Subjectivism and Global Structuralism prima facie fail to account for a sufficient condition insofar as they must be flanked by other physical and material factors, such as the ability of digesting the relevant food which are outside the domain of interpretation and human dependent structure. While we are not contending that a strategy for addressing these families of cases cannot in principle be developed by a subjectivist or a global structuralist, they are clearly facing substantial theoretical challenges.
We contend furthermore that not only Subjectivism and Global Structuralism, but also Local Structuralism lacks a straightforward strategy for accommodating our semantic practices and our ontological expectations with respect to what counts as food in all these cases. In fact, interpretation alone cannot turn a randomly taken natural thing into food-the agent should be endowed with relevant biological conditions for digesting it too. The argument can be generalized to additional and more puzzling cases too. Can an Amanita pantherina, a poisonous but not deadly fungus, be turned into food as long as someone interprets it as food? Should the general human conditions incapable of digesting it play a role beyond interpretation? 8 Negotiational Structuralism, instead, does have some resources to address the absentee enzyme test. In fact, it concedes that the interpreted things possess some independent features that posit certain constraints upon interpretations. These, therefore, should be aligned to the "one way" senses of reality that agents encounter. The specifics of the alignment, however, are open for negotiation based on different interpretations of what counts as food, such as: "milk is 8 Likewise, the premises of the argument above can explain how something becomes a food. For instance, according to an approximate historical reconstruction, three thousand years ago human beings realized that milk carried in bags derived from animals' stomachs can be stored slightly longer due to the presence of rennet, the enzyme that enables milk digestion, which curdles milk and starts the cheesemaking process (Dalby 2009: 30-35). Cheese was a perfect food for our ancestors since it lasted longer than milk and was always available, while being more digestible than milk itself. However, although it can be argued that acknowledging that cheese is food is an interpretational practice, this practice could not be carried out regardless of the ability to digest milk which was developed in association with the discovery, or invention, of cheese. In fact, dairy practice would have turned out to be useless and the interpretation vacuous if human beings lacked the ability to digest cheese. a staple food for our society and yet it cannot be food for me," or "milk can be food for me only if I take a pill," or "milk is food for me even if I'm lactose intolerant for I value its benefit more than its potential damages." All those interpretations rely on additional social conditions (e.g., the social and historical role of milk in many human diets, the distinction between firstperson and collective perspective, the attribution of value to a thing based on goals) and independent conditions (i.e., the presence of an enzyme, the digestive process, the effect of a pill on the body).

The Antagonistic Interpreter
The last test concerns the degree within which a theory can countenance different-sometimes inconsistent-interpretations of a given entity for the sake of respect and inclusion. Consider the following verisimilar scenario. An interreligious couple intends to pick a meat dish to serve at their wedding party. The religion of one of the partners regards pork meat as the most valuable food to be served on this special occasion and it regards cows as not food at all. The religion of the other partner regards beef as the most valuable food to be served on this special occasion and regards pigs as not food at all. At the same time, both religions require that there should be only one main dish for all the guests. Which meat should be served at the event? How can the disagreement be settled? 9 Negotiational Structuralism has the theoretical resources to address situations like the one just described, which call for a commitment to resolve a disagreement. In fact, even if a negotiational structuralist buys into a certain interpretation of what counts as food and what counts as the most appropriate food for a wedding, they also buy into the theoretical principle according to which such interpretations should be revised under new circumstances by taking into account non-interpretational fact as well as alternative interpretations.
Can Subjectivism, Global, and Local Structuralism solve the disagreement? If we stay with their core principles, we are offered no strategy to find a solution. Thus, a subjectivist can claim that, since each interpretation ultimately relies on the preferences and the background of the interpreter, there is no inherent reason that should push the members of the two religions toward a common solution. A global structuralist could claim that, as long as interpretations depend on other interpretations (all the way down), there is no reason in principle to reconcile the two religions. Local structuralists believe that there are as many ways of interpreting reality as there are possible interpretations of it, so that there is no principle that should encourage them to reach an agreement.
Of course, this does not mean that Subjectivism, Global, and Local Structuralism are incompatible with additional principles that would help to settle disagreements among antagonistic interpretations over what counts as food. Supporters of these three positions can invoke all sorts of additional ethical principles to supplement their views. Yet, it is remarkable that such positions do not bear a commitment to settling disagreements in and of themselves, so that there is no warranty that, by endorsing one of the positions, one also thereby endorses a commitment to try and bridge between opposing worldviews.

Conclusions
In this paper we elaborated on Kaplan's proposal that food is always open to interpretation. The frame we offered relies on conceptual tools proper of social ontology and analytic metaphysics. In doing so, we made an attempt at bridging an hermeneutic and an analytic perspective on the metaphysics of food in order to initiate what we think can be a fruitful dialogue.
We agree with Kaplan that interpretation plays an important role for providing conceptual aid to the confusions of everyday food ontologies and as a tool for social critique. In this sense, interpretation also serves to account for the dynamic and complex ontological nature of food. As we demonstrated, however, any view relying on interpretation must be supplemented with an articulated conceptual apparatus to explain challenging case studies. We also suggested that a promising avenue of research for the metaphysics of food explores further the negotiational structural reading, which more swiftly accounts for the collective processes of construction involving food identities. 10 The challenges we presented may require resorting to the existence of social and natural entities that are independent of specific acts of interpretation in some explanatory contexts. For instance, disparate entities such as basic biochemical processes, governmental agencies, or the built environment are-one may argue-key factors in shaping structures of interpretation that are not interpretations themselves. Thus, we end our discussion with a question that we cannot address in this paper: can interpretation alone suffice in founding the construction of food identities?
Funding Open Access funding provided by Università degli Studi di Milano within the CRUI-CARE Agreement. This research was founded by the Department of Philosophy "Piero Martinetti" of the University of Milan under the Project "Department of Excellence 2018-2022" awarded by the Ministry of Education, University and Research (MIUR).

Compliance with Ethical Standards
Conflict of Interest The authors declare that they have no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.