1 Introduction

The directedness of mental states, their aboutness, is their intentionality. For example, the belief that today is a fine day has an object—it is directed, it is about something, namely, today’s weather. It also has a subject—the believer. There are many accounts of intentionality; representationalism, inferentialism, and others. I’ll not address these, and I’ll not present an account of intentionality. Instead, as the paper proceeds I’ll mainly be concerned with the subject-side of intentionality, in order to address a debate in social ontology; namely, the debate about the relative fundamentality between collectivity and collective intentionality, on the one hand, and individuality and individual intentionality, on the other (a debate in which the many different accounts of the nature of intentionality are put to the side). This is the debate whether the ‘we’ or ‘us’ of collective intentionality—as when, e.g., we believe that today is a fine day—reduces to the ‘I’s or ‘me’s of individual intentionality—as when, e.g., you believe and I believe that today is a fine day and you believe and I believe that we both believe this. This is a debate about fundamentality.

What I call the fundamentality-question is which intentionality, individual or collective, is most fundamental. Is collective intentionality sui generis, not reducible to individual intentionality, yet caused and realized in individual brains (Searle, 1995, 2010)? Or is it rather reducible to the intentionality of individuals under conditions of common knowledge (e.g., Bratman 1992)?

Before looking at what others have had to say about the issue I want to introduce some terminology pertaining to the notion of fundamentality as I’ll address it later.

First, by some phenomenon A being fundamental for some other B is here meant that A cannot be accounted for in terms of B but B can be accounted for in terms of A. Second, by equifundamentality is meant that either both A and B can be accounted for in terms of each other, or neither can, and if there’s some other phenomenon, C, which either A or B can be accounted for in terms of then there’s some phenomenon (possibly the same, C) which the other (A or B) can be accounted for in terms of. Finally, that one can be accounted for in terms of another here means that either (a) the first doesn’t occur unless the latter does, or (b) one cannot grasp the concept for the first—know the circumstances of its proper application—unless one grasps the concept of the other. These two senses of account for will be further elaborated under the labels ‘reference-dependence’ and ‘sense-dependence’ below.

With that terminological clarification, the fundamentality-question has three answers, granting that equifundamentality is an option: either collectivity and collective intentionality or individuality and individual intentionality is fundamental, or they are equifundamental.Footnote 1

Prominent enquiries into the fundamentality-question include Gilbert (e.g., 1989; 2013), Searle (e.g., 1990; 1995; 2010), Bratman (e.g., 1992; 2014), Tuomela & Miller (1988) and Tuomela (e.g. 2005; 2013). More recently, Schmid (2018), Steiner (2012; 2014), and Gonzáles de Prado Salas and Zamora-Bonilla (2014; 2021) have made important contributions. For now, these will not be commented on. Suffice it to say that the orthodox view, not to say complete hegemony, is that, ontologically, there’s no irreducible collective subject of intentionality. Also someone who, like Searle, argues that the form or mode of collective intentionality (‘we intend’) doesn’t reduce to that of summed individual intentionality (‘I intend’ plus ‘I intend’) would be quick to add that that position doesn’t commit to an ontology of irreducible collective subjects. For no one, it seems, would want to be associated with the acceptance of, in Searle’s words, some “Hegelian world spirit” (1995: 25) or similar “abominable […] dreadful metaphysical excrescence” (1998: 150). Be that as it may,Footnote 2 the aim now is to fill a lacuna overlooked in previous contributions. What has been overlooked are different senses of fundamentality. When those senses are considered, it turns out that the fundamentality-question opens a richer field of possible answers than so far understood. Opening that field is the main purpose of this contribution.

This paper has two parts. The first is about the question itself. I aim to elaborate it (Sect. 2). According to the elaboration, there are two questions. First is the questions whether in either or both of the two senses given above (sense- and reference-dependence), either collectivity and collective intentionality or individuality and individual intentionality can be accounted for in terms of the other. As represented in Fig. 1, each sense of “account for” admits three answers (two for fundamentality, and one for equifundamentality). Combining the two questions (sense- and reference-dependence), nine answers are admitted.Footnote 3 In this part I also address two worries: one is that it is problematic to apply the distinction between sense- and reference-dependence to the issue at hand, and the second is that the distinction might be useful but that it fails to address the elephant in the fundamentality-room, namely, supervenience.

Fig. 1
figure 1

The elaborated Fundamentality-question

Full size image

The second part of the paper (Sect. 3) ventures a tentative answer to the elaborated fundamentality-question. To anticipate and simplify, the proposal is that in the order of reference-dependence there can be collectivity and collective intentionality only if there’s individuality and individual intentionality, but in the order of sense-dependence one cannot grasp what individuality and individual intentionality is unless one grasps what collectivity and collective intentionality is. The position is represented by combining cells 1 and 5 in Fig. 1. This is an anticipatory simplification complicated in the second part by considerations pertaining to an ambiguity concealed in the concept of individuality, and, correspondingly, in the concept of individual intentionality. Individuals can be conceived of as persons or sapients, on the one hand, and a sensuously aware organism, on the other. Arguably but tentatively, individuality and individual intentionality, in a robust sense, don’t have reference unless there’s collectivity and collective intentionality. That amounts to a equifundamentality answer (combining cells 3 and 6 in Fig. 1). To approach that answer, we need to elaborate the question.

2 Elaborating the Question: Sense- and Reference-Dependence

There are two ways to think about the relative fundamentality of two phenomena, which I introduced above. One is to think of the concept for one as having reference only if the concept for the other does. Brandom calls it reference-dependence,Footnote 4 saying that the “reference-dependence of one concept on another […] holds when the first cannot be true of something unless the second is true of something” (2015: 209).

A second way to think of relative fundamentality is of the grasp of one concept as depending on grasp of another. Thus, if one doesn’t grasp or understand the first concept, one doesn’t grasp or understand the other—the dependent. To grasp or understand a concept here means that one knows some of the conditions of its proper application and some of the consequences of its application. For instance, one doesn’t grasp the concept copper Footnote 5 unless one knows, e.g., that a condition for predicating it of some object is that the object conducts electricity and that a consequence is that the object would melt if heated to 1084.7 °C. Brandom calls this sense-dependence: “One concept is sense-dependent on another if one cannot grasp or understand the first without grasping or understanding the second” (2015: 209).

For purposes of illustration of reference-dependence, consider the concepts parent and child. It cannot be true of something that it is a parent unless it is true of something that it is a child, and vice versa. These concepts are equifundamental in reference-dependence. Also, one cannot grasp the concept parent without grasping the concept child. So the concepts are equifundamental in sense-dependence. For, a proper condition of application of child is that it has a parent, and a proper condition for applying parent is that it has a child.

I want to spell out some of the possibilities with respect to fundamentality-relations between concepts that the distinction between sense-and reference-dependence implies. This is important for the promised tentative answer to the fundamentality-question in a later section.

As the concepts female and sister help illustrate, two concepts may be relatively sense- and relatively reference-dependent. One cannot grasp the concept sister unless one grasps the concept female, because the proper conditions of application of sister contains as a condition that female applies. That is, anyone who grasps the concept sister already grasps, in principle,Footnote 6 the concept female. So, female is more fundamental than sister in sense-dependence. Also, there may be females but no sisters, but if there are sisters there are females. So female is more fundamental than sister also in reference-dependence. Of these concepts, one (female) is more fundamental in both senses.

It is trickier with mixed dependence. Mixed dependence obtains when one concepts is more fundamental in either sense- or reference-dependence, but the other more fundamental in the other, or one is more fundamental in one way but they are equifundamental in the other. Consider, for instance, copper and electrical conducer. A world without copper may still contain electrical conducers, but a world without electrical conducers doesn’t contain copper. If so, copper is reference-dependent on electrical conducer. But is the same true for sense-dependence? If not, their relation is one of mixed dependence. And it seems that there may be a world of concept-users who grasp the concept copper but not the concept electrical conducer. As a matter of fact, that situation once obtained in our world. That is, concept-users may know some of the proper conditions for applying copper but not for applying electrical conducer. So it seems that copper isn’t sense-dependent on electrical conducer. One may grasp the former without grasping the latter. Still, might it be that copper is sense-dependent on electrical conducer “in principle” in the way sister is on female (see footnote 6)? That is, might it be that the conditions for properly applying copper are implicitly known in knowing the conditions for properly applying electrical conducer, such that in possessing the concept copper one already implicitly knows everything one needs to know to possess the concept electrical conducer, whether one knows that or not? The answer is arguably negative. Additional conditions beside being an electrical conducer obtain for proper application of copper, such as pertain to, e.g., density and melting point. These conditions aren’t implicitly known in knowing proper conditions for applying electrical conducer. Electrical conducers come in a variety of materials the corresponding concepts for which have a variety of different conditions of proper application. Therefore, possessing concepts for materials which are electrical conducers isn’t, not even in principle, to possess the concept electrical conducer. That is to illustrate a case of mixed dependence. I’ve now given examples to illustrate the possibility of the nine combinations of fundamentality. That was an exercise to prepare for applying the distinction to the fundamentality-question. For ease of reference, I repeat Fig. 1 here.

I emphasise again that the main point with this paper is exactly to elaborate the fundamentality-question, as above. Doing so reveals dimensions of it that has so far not been addresses, and a richness of alternative answers thereby made available. I turn next to attempt to make sense of what others have had to say about the fundamentality-relationship between individual intentionality and the individuality it implies, on the one hand, and collective intentionality and the collectivity it implies, on the other, in both the order of sense- and reference-dependence.

As it turns out, and should be expected, the interpretative task quickly runs afoul because the elaborated fundamentality-question hasn’t been raised before. What others have said will therefore appear mediocre (not in the derogatory sense, but in the sense of lacking at least half of an answer to the question). So, if one accepts the distinction between sense- and reference-dependence as I draw it here, then the problem we face in making sense of what philosophers have said about fundamentality, or of what they should say about it given their other commitments, is a problem due not to the question I’m recommending us to pose but due rather to a nine-dimensional blind spot in answers so far received. Hence, before continuing, I’m obliged to justify the application of the distinction between sense- and reference-dependence.

As an anonymous reviewer pointed out, the distinction between sense- and reference-dependence, for Brandom, mainly applies to empirical concepts, while I’m applying it to a theoretical concept; namely, intentionality. Let me clarify and then answer.

Theoretical concepts, for Brandom, are those that apply only inferentially (e.g., 2015: 59–61). Consider the concept muon (standing for a posited sub-atomic particle). If we cannot directly observe muons but can only apply muon by inference from observed circumstances (hooked vapor trails in cloud chambers observed by properly trained particle physicists etc.), then muon is a theoretical concept (ibid.: 115–16). Empirical concepts, by contrast, are those we can apply noninferentially (ibid.: 114–15). For instance, red is an empirical concept. You don’t have to infer the application of red from the visual presence of a red thing.Footnote 7 That is the difference between theoretical and empirical concepts. That difference mustn’t be confused with the difference between theoretical and empirical objects. For Brandom, following Sellars, a theoretical object is one whose existence is inferentially posited. Consider the planet Neptune. The planetary body was first posited as a consequence of the mathematics that explained the movements of other planetary bodies. Thus it was a theoretical object. A few years later, aided by calculations and increasingly sophisticated telescopes, Neptune was observed. Now, an empirical object is an object that is observed. Neptune, then, was an empirical object; first it was inferentially posited (theoretical object) and then it was observed (empirical object). This isn’t a change of the object, in this case Neptune, but a change in how we’re acquainted with it (an epistemological and not metaphysical change) (Brandom, 2002: 362).Footnote 8

Now, then, the worry at hand is that intentionality is a theoretical concept and that the application of the distinction between sense- and reference-dependence is problematic in the context of theoretical concepts. But this worry is unwarranted. Thus, consider again the theoretical concept muon. We may ask, to begin with, what else would have to be true of a world for it to be true that it contains muons. That is to ask whether, and possibly on what, the theoretical concept muon is reference-dependent. It is an intelligible question. Also, answers to it make much sense: muon is reference-dependent on, e.g., mass and electric charge. We may also ask what other concepts anyone would have to master to master muon. This is to ask whether, and possibly on what other concepts, the theoretical concept muon is sense-dependent. It seems straightforward to search for an answer; e.g., one must grasp elementary particle, decay, and so on, to grasp muon. Hence, questions and answers about both the sense- and reference-dependence of theoretical concepts, of which muon is an example, can be shown to be unproblematic.

Recall that the worry that I want to answer here is that the application of the distinction between sense- and reference-dependence, as I apply it, is problematic because intentionality is a theoretical concept. But I’ve now shown an example of nonproblematic application of the distinction between sense- and reference-dependence to theoretical concepts. I’ll not go on to consider more cases, but assume the project worthy of pursuit.Footnote 9

We may now return to the interpretative task to find what others have said about fundamentality in light of the elaborated fundamentality-question. I said that this task is problematic because others haven’t considered that question before. Having now argued that the problem isn’t due to how I apply the distinction between sense- and reference-dependence in posing the question, I take it that the following shows why it is problematic to not apply that distinction in addressing the fundamentality-relations between the individuality and individual intentionality, on the one hand, and collectivity and collective intentionality, on the other. I will only consider two (rather different) cases: Searle, and Zahavi (and Zahavi and Kriegel).

Searle argues that collective intentionality is a sui generis form of intentionality, caused and realized in individual brains of members of sufficiently evolved species (1995: 24–25; 2006). This recommends interpreting him as claiming equifundamentality in reference-dependence. That is, if brains of members of those species cause individual intentionality, they cause collective intentionality, and vice versa (equifundamentality in reference-dependence). On the other hand, his position is perfectly compatible with there being worlds with no species having evolved collective intentionality but with some species having individual intentionality. So, if the reference-class is possible worlds rather than known species in our world, collective intentionality and individual intentionality aren’t equifundamental in reference-dependence. In fact, also if the reference-class is our world, Searle’s position is compatible with there being individual intentionality but no collective intentionality if he takes non-human organisms to have individual intentionality. Hence, it might still be that, for Searle, collective intentionality is reference-dependent on individual intentionality; i.e., in some, perhaps many, world(s) it is true of something that it has individual intentionality but not true that something has collective intentionality. But then again, his statement that no set of individual intentionality, in the form ‘I intend’, even supplemented with mutual belief, adds up to collective intentionality, in the form ‘we intend’ (1995: 24)Footnote 10 suggests the opposite interpretations. To correctly interpret Searle’s position in the realm of reference-dependence we need to know if it is compatible with the possibility of collective intentionality referring without individual intentionality referring.

Concerning sense-dependence it isn’t much clearer. Searle’s claim that in order to engage in collective agency, which involves collective intentionality, one need only to have a “Background sense of the other as a candidate in cooperative agency” (1990: 415) is suggestive. One needn’t, plausibly, have the concept collective intentionality to have collective intentionality; if there are (or are no) concept-users there can be collective intentionality without concept-users having the concept collective intentionality. The question then is if one must have the concept individual intentionality to have the concept collective intentionality, and/or vice versa, or if they are equifundamental in sense-dependence. That is, can anyone understand what collective intentionality is without understanding what individual intentionality is, and/or vice versa?

Since what Searle says isn’t a response to the elaborated fundamentality-question, no clear answer is to be expected. In fact, faced with this new question, Searle might give entirely different answers than the ones I’ve attributed to him. I’ll return to this later.

Zahavi and Kriegel are straightforward and explicit on reference-dependence, in the context of individual and collectively shared experience. They say that the first-person “me-ness-manner” of experiencing is a “pervasive” “constitutive” “feature of experiential life as such,” (2016: 38). It is “the most fundamental fact … most general, most elemental dimension about phenomenal consciousness” (ibid.: 50).Footnote 11 So, it would seem, there can be individual experiencing (and, I add, intentionality) without collective.

With respect to sense-dependence, matters are less clear. Granting both that there can be me-ness-manner experiencing without concept-users and that there can be me-mess-manner experiencing without us-ness-manner experiencing, we can still ask: could there be concept-users grasping the individuality implicit in for-me-ness experiencing that according to Zahavi and Kriegel is the most fundamental fact of phenomenal consciousness and which is expressed by the concept for-me-ness, or colloquially in ‘I feel …’, while those concept-users don’t grasp the collectivity implicit in shared for-us-ness experiencing, expressed by the concept for-us-ness, or colloquially ‘we feel…’? This is the question, not if there can be for-me-ness experiencing without for-us-ness experiencing, but if anyone could grasp and express what subjectivity is implicit in one of these modes of experiencing without also grasping and being able to express what subjectivity the other mode implicitly involves. Clearly, that isn’t Zahavi’s, or Zahavi and Kriegel’s, question, concerned, as they are, with phenomenal consciousness. But it is a question worth asking to properly appreciate the many facets that the fundamentality-question shelters, and to learn to navigate the kaleidoscope of answers that rotating them produces.

It is clear in much of the literature that reference-dependence is the main preoccupation; with whether collective and collective intentionality has reference only if individual and individual intentionality. The issue of sense-dependence has by and large passed unnoticed. I end this section with a note on why this neglect is important, using an analogy to how and why Brandom puts the distinction between sense- and reference-dependence to use. But first, I want to address those for whom all this talk of sense- and reference-dependence might be positively intriguing but ultimately an evasion of the real issue: supervenience.Footnote 12

Let us help ourselves to a standard and admittedly somewhat generic articulation of supervenience: properties of some type C´ supervene on some other type of properties C if it isn’t possible that two worlds are identical with respect to their C-type properties but not their C´-type properties; that is, the C-properties in a world determine its C´-properties. This is called global supervenience (e.g., Chalmers 1996: 34). In the present context, orthodoxy has it that collective intentionality supervenes on individual psychological states: if facts about psychological states of individuals in world w were different, so would facts about collective intentionality in w; or, if individuals in w didn’t have certain psychological states, w would lack collective intentionality. While these claims have an intuitive appeal, they aren’t uncontroversial (e.g., Epstein 2015). For purposes of this paper, I won’t dispute that collectivity and collective intentionality supervene on individuality and individual psychological states.Footnote 13

Still, someone could object that the sense of fundamentality that the concept of supervenience articulates is more fundamental than those articulated by sense- and reference-dependence, respectively or together. This is a critique directed against the relevance of working with the latter two. I cannot devote much space to a detailed argument whether supervenience is a more fundamental fundamentality-relation than are sense- and reference-dependence, let alone give a satisfyingly exhaustive analysis of how it relates to the latter two.Footnote 14 Nevertheless, note that supervenience and reference-dependence aren’t worlds apart. After all, while reference-dependence, on the one hand, articulates a relation of fundamentality between concepts C and C´ such that “C´ is reference-dependence on C” means that unless C is true of something in world w, C´ isn’t true of something in w, supervenience, on the other, articulates a relation of fundamentality between properties, say C-type and C´-type properties, such that “C´ supervenes on C” means that there are no two possible worlds w and w´ identical with respect C-properties but differing with respect to C´-properties. But, now, if concept C´ (say, child) is reference-dependence on concept C (say, parent) then it is also true that there would be no C´-property (property picked out by C´; say, child of) in w if there were no C-property (property picked out by C; say, parent of) in w. What reference-dependence doesn’t imply, while supervenience does imply it, is that C-facts determine C´-facts (locally or globally). What reference-dependence implies, rather, is that whether C refers in w determines (I prefer “settles the question”) whether C´ can refer in w; or, in other words, the fact whether C refers in w determines (settles) the fact whether C´ can refer in w. We could perhaps say that supervenience helps in explaining why worlds differ or are identical with respect to certain properties and facts. Reference-dependence instead helps to determine what judgements about properties and facts can be true in a world depending on what others are.

What I’ve said doesn’t answer a critic who objects that supervenience is a more fundamental fundamentality-relation than sense- and reference-dependence. Instead I’ve considered similarities and differences between supervenience and reference-dependence. Sense-dependence, I hope I’ve made clear, is a different beast. Which naturally leads to the closing of this section.

I said I’ll close this section by clarifying why sense-dependence is important, using an analogy. Here it is: Brandom introduces the distinction between sense- and reference-dependence to defend the dual and according to him complementary theses of modal realism and modal expressivism (2015, Chapter 5; 2019, Chapter 7). Roughly, he argues that true modal claims state facts (modal realism), and that what one is doing in making modal claims is making explicit inferential relations implicit in making ordinary empirical descriptive claims such that the latter track modal facts (modal expressivism). The ultimate aim is to be entitled to say that the classical empiricist Humean, or more updated Quinean, position, in which one purports to understand perfectly well statements to the effect that things are thus-and-so but not having a clue on what grounds statements are warranted that they could or must be some way, are mistaken (Brandom, 2015: 179, 213). To that end, the argument is that expressions as to how things are (not) are sense-dependent on expressions as to how things could (not) be. That is, one cannot grasp ordinary empirical descriptive expressions, e.g., “The cat is on the mat,” unless one grasps their modal involvements, e.g., that it could still be true if the cat and the mat were moved slightly to the right or if a bumblebee in the garden were about to pollinate a daffodil but wouldn’t be true if the mat burst into flames or the cat ran off to chase a mouse. This would mean that one must already be able implicitly to grasp modal claims to grasp ordinary descriptive claims.

The analogue I want to make is that adding sense-dependence as a dimension of concern along common concerns with reference-dependence is important because in so doing we may ask whether, assuming we grasp individual intentionality and what it is for an ‘I’ or ‘me’ to have intentionality, we must already grasp collective intentionality and what it is for a ‘we’ or ‘us’ to have intentionality, and/or vice versa.

Thus, imagine the claim, on analogy to the Humean’s about laws and necessities, that one can understand perfectly well what individuality and individual intentionality is but not what collectivity and collective intentionality is. Can we answer, analogously to Brandom’s response to the Humean, that in knowing the former one already knows everything one needs to know to know the latter? Continuing the analogy (perhaps stretching it too far), can we have the combined and complementary theses that, first, claims about collectivity and collective intentionality state facts that don’t reduce to facts about individuality and individual intentionality (realism about collectivity and collective intentionality) and that those claims make explicit capacities implicit in making claims about individuality and individual intentionality (expressivism about collectivity and collective intentionality)? Equipped with the distinction between sense- and reference-dependence, the project to assess whether we can be to Searle, and others (or indeed to the major part of the received wisdom on the fundamentality-question), what Brandom purports to be to Hume and Quine, and others, can start. What follows in the next section is a first attempted ant step on that path.

And let me add another case in which the question of sense-dependence is interesting to raise. Earlier we considered a critic asking if not supervenience articulates a more fundamental sense of fundamentality than do sense- and reference-dependence. I admitted that I couldn’t (at least here) give a straightforward answer. But now consider this counter: Are you asking if not supervenience gives a more fundamental sense of fundamentality than either sense-dependence or reference-dependence, or both, in either sense-dependence or reference-dependence, or both, or a more fundamental sense than one in one sense but equifundamental to the other in the other sense? Perhaps it is intelligible to ask (but, at the face of it, not to me) whether supervenience refers to something only if either or both reference-dependence or sense-dependence refers to something, or the reverse, or if they’re equifundamental in reference-dependence. One could also ask whether one has to grasp supervenience in order to grasp either or both reference-dependence or sense-dependence, or the reverse, or if they’re equifundamental in sense-dependence. But I must admit that for theoretical concept of this calibre I stand dumbfound at the question; the question, namely, if not supervenience is more fundamental.Footnote 15

I said that the elaboration of the fundamentality-question was my first and main concern. So the paper might’ve ended here. But I want to venture one more tempting and natural idea; namely, a tentative answer.

3 Further Elaboration and a Tentative Answer

The first part of the answer I now propose is that one cannot grasp what individuality and individual intentionality is without also grasping what collectivity and collective intentionality is. This is to claim the fundamentality of the second pair, in the realm of sense-dependence. The second part of the answer is that unless there’s individuality and individual intentionality there’s no collectivity and collective intentionality, respectively. This is to claim the fundamentality of the first pair, in the realm of reference-dependence. In the end, though, I find reasons for a stronger, incredulous stare-provoking claim that, in one sense, there also wouldn’t be individual intentionality and individuality in a world without collective intentionality and collectivity.

The place to start is with a well-known ambiguity in the concept individual, which affects also how to think of individual intentionality. The ambiguity is brought out by the question whether, to paraphrase Sellars (1962: 76–77), there’s a difference between how we think of featherless bipeds, and the way they have intentionality, if they do, on the one hand, and persons, and the way they have intentionality, on the other. Persons, for Sellars, contrast with organisms in that they can enter relations of right and duties, entitlements and commitments. Persons hold and can be held responsible. That is a relation of mutual normative recognition, which is the institution of community—not in the metaphysically loaded sense of one mind or body but in the deontological sense of sharing norms, standards and principles (Brandom, 2004; Steiner, 2014). The sharing of standards and principles is what makes discursiveness and rationality possible, for Sellars; members of community can pursue the game of giving and asking for reasons, of demanding, producing and consuming justifications for what they say or otherwise do (Sellars, 1956: §36; cf. Brandom, 2009: 128). Individuals who can do that, i.e., persons, are ‘I’s in a rather different sense from featherless bipeds. And individuals in that sense have individual intentionality in a rather different sense than featherless bipeds do, if they do. While individuals in the persons-sense can ask for and give reasons for believing, acting and intending one way rather than another, individuals in the organism-sense reliably respond differentially to stimuli but do not also reason about reliability, when to respond how and why, and so on. This, then, suggests two different senses of both individual intentionality and of the individuality implied. Sellars says of the community, the ‘we’ thus understood, that it is no less basic than persons (1962: 77). What he means, I take it, is that individuals, in the sense of persons, are not possible in abstraction from a community within which alone the game of giving and asking for reasons becomes possible. At this point it ought to come as no surprise that I take Sellars to make a case for the claim that although person and individual intentionality (of persons) are reference-dependent on organism and individual intentionality (of organisms), respectively, the first pair is also reference-dependent on collective intentionality and the we with which it is associated, respectively.

In a contemporary context, Sellars’s conceptual framework of persons corresponds to Brandom’s conceptual framework of sapience, someones or subjects, as opposed to sentience, somethings or objects (2000: 81, 157–58; 2009: 3, 10, 135). It also reminds of what, in a different vein, Baker refers to as persons with robust first-person perspectives, as opposed to organisms with rudimentary first-person perspectives, which are constitutive of but not identical to persons (2000, Chapter 6; 2015).

Let us look closer at the supposed difference the concepts mentioned convey. Let us say that sentient individual intentionality is what organisms exhibit in capacities to reliably respond differentially according to needs and stimuli. It is, for instance, exhibited by cats chasing mice and by mice chased by cats. I think it is what Searle has in mind under the label ‘consciousness’, when he says.

consciousness consists of those states of feeling, sentience, or awareness that typically begin when we wake from a dreamless sleep and continue throughout the day until […] we go to sleep, go into a coma, die, or otherwise become “unconscious”. (2005: 10)

The concept sentient individual intentionality is, obviously, reference-dependent on there being creatures with the relevant capacities for reliable differential responsiveness, feelings, and so on. And, clearly, there may be sentient individual intentionality, and sentient individuals, in a world without concept-users. As far as we know, this was true of our world until human organisms developed sapient capacities.

Here is what Brandom says of the difference between sentience and sapience:

If we look just a little bit closer at our consciousness or mindedness, we see that it comes in two importantly different flavors. We are sentient beings, and we are sapient beings—we feel, and we think. Sentience is sensuous awareness, of the generic sort exhibited by at least our mammalian cousins. Paradigmatic states of sentience are feeling pain, seeing colors, and hearing sounds. Sentient awareness is what an organism has when it is awake, but, dreaming aside, not when it is asleep. Sapience is conceptual awareness—a kind of mindedness that is tied to understanding rather than sensing. Paradigmatic sapient states are thinking or believing that things are thus-and-so (or desiring or intending that they be thus-and-so) (2009: 135).

So, sapient individuals can, whereas non-sapient sentient individuals cannot, reason about how to act and about what is a reliable, successful, correct or appropriate response to, and action in, some contingency. Importantly, sapient individual intentionality is significantly different from its sentient cousin, if it has such a cousin, in that it involves capacities to believe, desire and intend that thus-and-so.

Finally, Baker approaches the concept person as follows:

Only persons can have fears and hopes about the future, and only persons can attempt to shape their futures according to their own ideas of the kinds of beings that they want to be. In short, animals that do not constitute persons cannot be important to themselves in the same way that persons are important to themselves. This is so because the first-person perspective allows us to think about, and conceive of, ourselves in a unique way […] Persons are important in the scheme of things as bearers of normativity. […] A person that torments babies is reprehensible; a cat that torments mice is not. (2000: 147)Footnote 16

Persons are distinct, for Baker, in that they “not only have goals and pursue them” but are also able to evaluate their goals. Persons can, whereas non-persons cannot, “ask ‘Is this a goal that I really should have?’” (ibid.: 148).

Now, if we take seriously the differences between conceptual frameworks—between individuality in the normative sense of persons or sapients, on the one hand, and organisms or sentients, on the other, and between conceptual or sapient individual intentionality, on the one hand, and sentient or organismic individual intentionality, on the other—then the fundamentality-question turns out to shelter yet another dimension of possible answers. For we can now ask about the sense- and reference-dependence between these new senses of individuality and individual intentionality.Footnote 17 So, let me close by considering what I above called the strong, incredulous stare-provoking claim; the claim, namely, that individual intentionality and the individuality implied, on the one hand, and collective intentionality and the collectivity implied, on the other, are equifundamental in reference-dependence, while the latter are fundamental in sense-dependence. The felt incredulity is mitigated, I argue, if we keep firmly in mind what sense of individuality is intended, and what sense of individual intentionality.

To begin with, despite important differences it is reasonable to assume a tight coupling between sentience and sapience (Baker, 2005, 2015; Brandom, 2009). It is difficult to think of non-sentient sapients. Such beings wouldn’t be embodied, sensing or situated but nevertheless be capable of giving, asking and assessing reasons for and against beliefs, desires, and ways of acting. Perhaps a godlike being could do so. Bracketing such possibilities, in what follows I’ll consider only alternatives where sapient individual intentionality is reference-dependent on sentient individual intentionality.

The reference-dependence of sapient individual intentionality on sentient individual intentionality is compatible with the latter being sense-dependent on the former. For consider a world with sentients but no sapients; there are no concept-users, hence no one grasping or understanding either sentient individual intentionality or sapient individual intentionality. Put otherwise, since only sapients are concept-users, concept is reference-dependence on sapient individuality. This entails that the concepts sentient individual and sentient individual intentionality are reference-dependent on sapient individuality.Footnote 18 This doesn’t entail the reference-dependence of sentient individuality or sentient individual intentionality on sapient individuality or sapient individual intentionality. It entails that one can grasp the former concepts—indeed any concept—only if one is sapient; hence, also that there are concepts only if there are sapients.

Now if individuality in the sense of sapience is only possible within the context of a community with principles and standards that make giving and asking for reasons possible—a normative context of mutual recognition—then sapient individuality, hence also sapient individual intentionality, is reference-dependent on there being some such community. On this view, sentient individuals with their sentient individual intentionality achieve sapient individuality and sapient individual intentionality only together. The we, and the collective intentionality associated with it is then, to paraphrase Sellars (1962: 77), no less basic than the individuals that make it up. But it is of utmost importance that the individuality conveyed by “individuals” is here meant in the sense of sapient individuality! Sellars is speaking of individual persons, which for him means individuals with socially articulated normative statuses (sapients), not of featherless bipeds, which for him means the organismic nature of persons and non-persons (sentients). If we keep that firmly in mind, then the claim that individual intentionality and individuality are reference-dependent on collective intentionality and collectivity ought not cause the ruckus it typically does. It is only if you mistake that claim for the claim that there couldn’t be organisms and sentient intentionality in abstraction from collectives and collective intentionality that an incredulous stare is justified. But it is a mistake that, according to my argument, is traceable, first, to a failure to disambiguate individuality, hence also individual intentionality, and, secondly, to a failure to disambiguate fundamentality (sense- and reference-dependence). Together, those failures hold one hostage to the framework of the unelaborated fundamentality-question, from which the incredulous stare-interpretation of the fundamentality of collectivity and collective intentionality appears to be the only one possible. Then one might end up talking pejoratively about a “Hegelian world spirit” (Searle, 1995: 25) or other “abominable […] dreadful metaphysical excrescence” (Searle, 1998: 150).Footnote 19

So far, I’ve argued for the reference-dependence of sapient individual intentionality, hence also of sapient individuality, on both sentient individual intentionality, hence also on sentient individuality, and on collective intentionality, hence also on collectivity (the we or community within which giving and asking for reasons can be pursued as a collective practice). At the same time, I’ve admitted the reference-dependence of collective intentionality, hence also of collectivity, on sentient individual intentionality, hence also on sentient individuality. As far as reference-dependence is concerned, we may conclude that in one sense of individuality and individual intentionality, namely sapience, collectivity and collective intentionality are more fundamental, while in another sense, namely sentience, individuality and individual intentionality is more fundamental.

In what remains I turn to sense-dependence; that is, to the fundamentality-relations between possession of the concepts sentient individual intentionality and sentient individuality, sapient individual intentionality and sapient individuality, and collective intentionality and collectivity. The question is if one can grasp any of these concepts and, if so, which, without grasping others and, if so, which.

Suppose to begin with that we grasp sapient individuality and sapient individual intentionality. On the present approach, we take it to be a capacity to pursue the game of giving and asking for reasons in community with others with whom we articulate what counts as a reason for or against what. The question is whether we then already grasp either or both (a) sentient individuality and sentient individual intentionality, which, on the present approach, is to be an individual capable to reliably respond differentially to stimuli as exhibited in behavioural discriminations, or (b) collectivity and collective intentionality, which on the present approach is to be a collection of individuals capable to together articulate what counts as a reason for or against what, such that they can and do assess under what circumstances it is proper to act or judge one way rather than another. To be exact, the question is whether sapient individuality and sapient individual intentionality is sense-dependent on either (a) sentient individuality and sentient individual intentionality and (b) collective and collective intentionality, or on neither or on both. And note here the ambiguity of “collection of individuals” in the first formulation of (b), implicit in collective in the second formulation. This is the first hurdle (anticipated in footnote 15 above). For what sense of individuality is implicit in collection of individuals, which in turn is implicit in collective, makes all the difference for what we go on to take ourselves to understand by community or we and collective intentionality, which makes all the difference for the whole issue of sense-dependence.

I take it to be straightforward that the sense of individuality implicit in collection of individuals of a community or we must be sapient individuality. The reason why is that if it were sentient individuality then we’d be speaking about reliable dispositions to respond differentially and so on, and not capacities to articulate reasons and pursue the game of giving and asking for reasons. Perhaps we could make sense of a notion of sentient community or sentient we; a collective of sentient individuals—the behaviour of gregarious animals and social insects comes to mind.Footnote 20 But if by collective intentionality we think of intentionality expressible by “We intend to…”, “We believe that…”, and so on—and this is indeed the standard construal in the literature on collectivity and collective intentionality—then it is highly implausible that the sense of individuality implicit in collection of individuals of a we or community should be sentient individuality, i.e., intentionality exhibited in capacities to reliably respond differentially to stimuli. The consequence of this reasoning is that sapient individuality and sapient individual intentionality is sense-dependent on collective and collective intentionality. For according to the argument, one cannot grasp the former without grasping the latter. One must grasp the concept collectivity and the related collective intentionality (in the sense of a collection of sapient individuals) to grasp the concept sapient individuality and related sapient individual intentionality.

I now claim, further, that one cannot grasp sentience, sentient individuality or sentient individual intentionality unless one, at least in principle (see footnote 6 above), already grasps sapience, sapient individuality and sapient individual intentionality. That is, the former are sense-dependent on the latter. Admittedly, the reasoning in favour of that claim is somewhat speculative. Nevertheless, consider a community of sapients. On the present proposal, they treat each other as and they are capable to give and ask for reasons, they are and treat each other as beings with rights and duties, who can take responsibility and hold and be held responsible. In Sellars’s terms, they treat each other as persons. In Brandom’s, they take each other to be someones (as opposed to somethings). In Baker’s, they have robust first-person perspectives. What they thereby implicitly do—i.e., whether or not they can articulate that they do it—is grasp sapience, sapient individuality and sapient individual intentionality (though it may be that they don’t use those or other concepts (who does?)).Footnote 21 This doing—that is, treating something as x—is what having a capability of grasping x “in principle” means. Something our imaginary community members do not thereby necessarily already know, not even in principle, is what it is for others or themselves to be sentient organisms. We have a living example of a species whose members didn’t grasp, not even in principle, their organismic nature before they treated each other as persons: us, homo sapience. This is another Sellarsian point: it isn’t plausible to suppose that our first human encounters with ourselves required us to grasp that we were complex biological compounds (sentients), which we then had to learn to treat as persons (sapients). Quite the opposite: grasping our sentient organismic nature is a late-won scientific achievement. Grasping sentient individuality and sentient individual intentionality is an achievement which required an antecedent in principle grasp of sapient individuality and sapient individual intentionality. Speculative as the route to it might be,Footnote 22 it is plausible, then, that sentient individuality and sentient individual intentionality are sense-dependent on sapient individuality and sapient individual intentionality.

To conclude, I’ve already argued that sapient individuality and sapient individual intentionality are sense-dependent on collectivity and collective intentionality. Together with the argument in the previous paragraph, that sentient individuality and sentient individual intentionality are sense-dependent on sapient individuality and sapient individual intentionality, the conclusion is that sentient individuality and sentient individual intentionality are sense-dependent on collectivity and collective intentionality. Before that, I argued that though sapient individuality and sapient individual intentionality are reference-dependent on collectivity and collective intentionality, the latter pair is reference-dependent on sentient individuality and sentient individual intentionality.

So, in summary, orthodoxy is correct: there’s no collective or we with collective intentionality if there aren’t individuals with individual intentionality, in the sentient sense of individuality and intentionality. Who would dispute that? Still, unless there’s some collective or we with collective intentionality, there aren’t individuals or individual intentionality, in the sapient sense of individuality and intentionality. And what is more is that one cannot grasp what sentient or sapient individuality and individual intentionality is unless one grasps what collectivity and collective intentionality is. Hence, I conclude with the following combination of claims:

  1. (a)

    sentient individuality and sentient individual intentionality is fundamental in reference-dependence across the board,

  2. (b)

    sapient individuality and sapient individual intentionality is reference-dependent on collectivity and collective intentionality, and.

  3. (c)

    collectivity and collective intentionality is fundamental in sense-dependence across the board.

The following table summarises the conclusion, with cells ordered in the two dimensions of fundamentality where upper position indicates dependence on the lower one(s) (Fig. 2):

Fig. 2
figure 2

A tentative answer to the elaborated fundamentality-question

Full size image