Human-Relative Values as Epistemic Challenge in Heraclitus

1 Introduction

In his human-animal juxtapositions, Heraclitus examines the different qualities that objects seem to have when considered relative to the natural needs and preferences of different creatures.^[1] In B13, a human-relative conceptions of “pure water” (καθαρῷ ὕδατι) and of “mire” or “filth” (βορβόρῳ) are juxtaposed with a pig-relative conception of the same^[2]; in B9, a human-relative conception of refuse (σύρματα) is contrasted with a donkey-relative conception of gold (χρυσός); and in B61, a human-relative conception of seawater is contrasted with a fish-relative conception. One central interpretation of these fragments takes them as attacking human values by the paralleling their sources and paradigmatic objects with those of nonhuman animals.^[3] I will call this the ‘values reading.’ This reading is made more plausible by other fragments that draw upon the symbolic associations of nonhuman animals in complaining of the ignorance of some subset of humanity.^[4] A second interpretation, which I will call the ‘unity reading,’ takes these fragments as instances of the unity of opposites.^[5] On this reading, the donkey-relative conception of refuse as choiceworthy and the human-relative conception of refuse as worthless provide a pair of opposites (choiceworthy, worthless) unified through their compresence within the refuse itself. This reading is made more plausible by similarities to other fragments that offer uncontroversial instances of the unity of opposites through their compresence in a particular object or variety of object.^[6]

These two readings, which are generally regarded as incompatible, diverge with respect to their understanding of the relatives by which the opposites are qualified. The unity reading allows both the donkey-relative conception and the human-relative conception some degree of insight into the object; the two conceptions yield opposites that, taken together, are constitutive of the object’s nature.^[7] In other words, the unity reading moves from two relative claims about the choices of both donkeys and human to a non-relative insight into the objects of choice. On the values reading, by contrast, humans start off by assuming a non-relative insight into the object, which Heraclitus juxtaposes with a second, evidently relative, claim about the needs and preferences of nonhuman animals. The parallelism between the two claims reveals that both are merely an expression of the natures of the species in question. As such, neither one gains traction on the nature of the object as such. Each creature’s object of choice might be taken as ‘choiceworthy’ for it, since refuse and gold support the survival of donkeys and humans respectively (by acting as nourishment or as a means of obtaining nourishment).^[8] Yet the apparent value of these ‘goods,’ on the values reading, is a function of the natures of creatures who depend for their survival upon what is actually ‘base’ and who, therefore, have an inborn fault through which they fail to discern what is really valuable. Thus, what the unity reading takes as an alignment and unification of the opposites ‘choiceworthy’ and ‘worthless’ within garbage and within gold, the values reading regards as an alignment and ‘unification’ of the human and the donkey – at least with respect to the two creatures’ equal inability to recognize and pursue real value.^[9]

In this paper, I suggest that Heraclitus’ human-animal juxtapositions call upon the audience to apply both the values reading and unity reading, yet ultimately leads them beyond both. Specifically, I argue that the juxtapositions use the challenge to human-relative values to support an even more fundamental challenge to the faculties by which we identify and distinguish objects. These basic distinctions, it turns out, cannot be separated from human evaluative discernment, and, as a consequence, are undermined along with the latter.

I argue that these juxtapositions, in emphasizing the idiosyncratically human evaluative ground of our conception of various objects, raise a series of deepening puzzles about our epistemic capacities. These fragments do, as the values reading contends, constitute an attack on human discernment of evaluative or axiological qualities, placing the audience on the same footing as animals traditionally associated with ignorance and ignobility. Yet the full force of the juxtapositions can be felt only by an audience that treats them as genuine instances of the unity of opposites. In fact, understood in relation to the unity reading, the challenge to human discernment in the values reading has an even greater scope than its name, and its usual application, brings out. The animal-human juxtapositions do more than simply problematize human values. Rather, in emphasizing the role that purportedly inborne or ‘natural’ human preferences play in human values, and the role that human values play in structuring oppositions and in picking out and differentiating various objects, these juxtapositions raise fundamental doubts about the adequacy of our faculties for comprehending the true character of things.

My interpretation in this article is shaped by two methodological starting-points. First, Heraclitus’ sayings are deliberately structured in such a way as to contain multiple layers for the audience to unpack. Second, Heraclitus has a philosophical aim in mind; that is, there is more at stake here than the play of possible meanings. Taken together, these two starting-points imply that the meanings are also intended to be, to some degree, sequential for the audience. This is a methodological assumption for a range of different philosophical approaches to Heraclitus, including, for example, the interpretations of the sayings as mini-inferences or arguments with an implicit premise or implicit conclusion, allegorical images, or vehicles of experiential insight.^[10] On such interpretations, the sayings are both multi-levelled and targeted; the more immediate, initial levels of meaning open onto the less obvious ones. For instance, Kahn 1979 and McCabe 2015 have argued that Heraclitus plays upon commonsense certainties through the use of concrete examples drawn from daily life (McCabe) or evokes and frustrates formal and semantic expectations drawn from the literary tradition (Kahn) in order to hold up a mirror to the audience’s own assumptions and expectations, while also problematizing them.

Throughout this article, I speak of the responses of ‘Heraclitus’ audience.’ In doing so, I emphasize the familiar pragmatic, literary and conceptual reference points of the late Archaic Ionian, as distinct from the responses and reference points of later readers. In other words, the phrase is intended to draw attention to the communicative context that would have shaped the range of ways in which Heraclitus might hope to express a given meaning. In particular, I hope to draw attention to cases where the most immediate or ‘obvious’ reference points for the audience might differ from those of the twenty-first century reader. The audience to which this article refers is, however, always an idealized audience; my primary aim is to interpret Heraclitus, not his reception.

Throughout this article, I use the term ‘natures’ to refer either to the general character of a species (e.g., donkey, pig, fish) or to that of some particular kind of ‘stuff’ or object (e.g., gold, water, refuse). This is meant to mark out the cluster of features through which the audience makes category distinctions. The belief that such natures existed would have been present in Heraclitus’ audience as a pre-reflective starting-point for Heraclitus to take up, exploit and – on my argument – problematize.

A deterministic conception of natures in the case of nonhuman animals is well-attested in the literature for this period – in other words, the assumption that being a specific kind of creature dictated a specific, limited range of responses.^[11] With respect to human nature, on the other hand, this sort of deterministic attitude seems to be specifically associated with considerations of humans under the aspect of ‘mortal thing’ – and, consequentially, with discussions that emphasize the frailty and limitations of the human species.^[12] This conception of the natures of both nonhuman animals and humans qua mortals might be analyzed in terms of a creature’s physiologically-grounded needs, preferences and need-fulfilling capacities. This conception of creaturely nature need not, however, imply a soul-body distinction, since here, the conception of mortal limits yields the necessary elements of the ‘physiologically-grounded’: the conditions of survival (or death) and health (or sickness) that ‘dictate’ certain patterns of choice or avoidance.^[13]

With respect to object natures, the importance of what I call the audience’s symbolic starting-points (see section 3 below) in framing their categories attests to presence of a pre-reflective assumption that natures in this sense existed. It also emphasizes the fact that such a starting-point need not imply theoretical baggage around definitions and essential traits. For instance, if one of the qualities through which I identify water is that I can swallow it and splash around in it, then I can determine that humid air is not water without ever formulating the relevant cluster of qualities through which I demarcate that category. In the animal-human juxtapositions, Heraclitus exploits the pre-reflective assumption that the audience knows such everyday objects as water, refuse and gold in this sense – as well as the assumption that the failure to know such basic objects represents the nadir of ignorance. Yet he also draws upon the expectation, set in place by his own illustrations of the unity of opposites as found in paths, rivers and carding-combs, that the knowledge of such objects may require a recognition of the unity of opposites within them.^[14]

I begin my argument in section 2, by introducing the juxtapositions and showing how Heraclitus uses species-relative qualifiers to both create and frustrate the impetus towards the unity reading. I also consider some of the problems that these creature-relative qualifiers present in the context of the late Archaic conceptual framework, especially considered against their re-framings in later philosophical tradition. In section 3, I argue that the attempt to apply the structure of the unity reading to the creature-relative opposites leads the audience to the values reading. This creates the first level of epistemic challenges, which concerns the value and disvalue, or axiological qualities, of objects. In B9 and B13, specifically, I argue that the juxtapositions are framed in such a way that the axiological opposites remain stuck within the creature-relative qualifiers, raising the worry that both the human-relative and the animal-relative axiological opposites might simply be an expression of idiosyncratic creaturely natures, rather than of the nature of the common object.

Section 4 shows how the human-fish juxtaposition in B61 deepens the first layer of epistemic challenges. In this fragment, even apparently non-axiological opposites come apart from the object; these, too, turn out to be reducible to creature-relative assessments of value and disvalue. Section 5 returns to B9 and B13, arguing that, once this additional epistemic challenge is recognized, these fragments can be seen to raise further questions about whether the distinctions that humans make between objects are a function of idiosyncratic creaturely natures. Section 6 enlists the aid of a final human-animal juxtaposition, B4, in pointing forward to some unanswered questions about the relationship between creature-relativity and language.

2 Pragmatic Assumptions: Motivating the Unity Reading

In this section, I introduce the human-animal juxtapositions and argue that each is structured in such a way as to invite the unity reading from the audience. Specifically, each contains an animal-relative conception of an object or objects, contrasted against a human-relative conception of the same. I argue that the audience’s basic pragmatic starting-points furnish them with good reasons to accept each of the two creature-relative statements brought together in any given juxtaposition. More precisely, these provide the common basis according to which the statements might seem to be both true and creature-relative. Both statements are creature-relative in the sense that they are grounded in the natures of the creatures in question, where the latter is understood as the creature’s basic physiological conditions of survival or well-being. Yet the plausibility of the two claims, considered singly, motivates the view that they refer to traits really possessed by the object. This view, in turn, generates a movement towards the unity reading – without, however, indicating how the audience is to move from the creature-relative qualifiers to the object.

I take Heraclitus’ human-animal juxtapositions to include the comparisons to fish, donkeys and pigs in fragments B9, B13 and B61.^[15] These three fragments run as follows:

B9: “Donkeys would choose refuse rather than gold.” (ὄνους σύρματ’ ἂν ἑλέσθαι μᾶλλον ἢ χρυσόν.)

B13: “Pigs enjoy mire rather than pure water.” (ὕες βορβόρῳ ἥδονται μᾶλλον ἢ καθαρῷ ὕδατι.)

B61: “The sea is the purest and most polluted water: drinkable and life-preserving for fish, but undrinkable and deadly for humans.” (θάλασσα, ὕδωρ καθαρώτατον καὶ μιαρώτατον, ἰχθύσι μὲν πότιμον καὶ σωτήριον, ἀνθρώποις δὲ ἄποτον καὶ ὀλέθριον.)

The main clause of B61 offers what seems to be an explicit expression of the unity of opposites: “The sea is the purest and most polluted water.” At the same time, the dependent clauses give us an explicit expression of the creature-relative conceptions, qualifying the opposites “pure” and “polluted” through datives of relation^[16] that identify the creatures relative to whose needs and preferences seawater might be said to possess these properties (seawater is pure for a fish; seawater is polluted for a human).^[17]

Yet, as widely observed both by commentators who favor the values reading and those who favor the unity reading, the presence of an animal-relative conception and a human-relative conception is clear in B9 and B13 as well. Like B61, these fragments explicitly name the creature for whom refuse is more choiceworthy than gold (B9) or mire more pleasant than pure water (B13). The fact that B13 and B9 invoke not only a valuation relative to the creature in question, but a contrasting, human-relative valuation is evident in the fact that the animal-relative valuation at once invokes and inverts human standards of the pleasant or choiceworthy. It is not simply the fact that the human preferences are the converse of the animal preferences in these instances – or even the fact that this is what makes the animal preferences worth remarking at all. In both cases, the objects occupy opposite poles within a human value scheme – with “pure water” and “gold” functioning almost as metonyms for value and with “filth” and “refuse” functioning as metonyms for worthlessness or disvalue. Thus, the human-relative conception is also implicitly present in the sense that the relevant objects actually express the two opposites, value and worthlessness – and do so from within the human-relative conception.^[18] The phrase μᾶλλον ἢ (rather than) used here has exclusive force^[19]; this is what makes these juxtapositions analyzable as instances of the unity of opposites. The exclusive force lies in the polarity between value and disvalue – like the objects themselves, the juxtapositions create an opposition between the choiceworthy and the worthless, the pleasant and the unpleasant that cuts along creature-relative lines.

B9 and B13, then, function via a juxtaposition of two creature-relative claims:

B9: “Donkeys would choose refuse rather than gold” (or, “For donkeys, refuse is choiceworthy [but not gold].”)

B9^a: “Humans would choose gold rather than refuse” (or, “For humans, gold is choiceworthy [but not refuse].”)

B13: “Pigs enjoy mire rather than pure water” (or, “For pigs, mire is enjoyable [but not pure water].”)

B13^b: “Humans enjoy pure water more than mire” (or, “For humans, pure water is enjoyable [but not mire].”)

The audience can affirm these separate statements not only on the basis of observations about what the creature tends to do or choose, but also on the basis of what is actually good for each creature in the sense of supporting its survival or flourishing. In other words, each statement appeals to a body of pragmatic beliefs about the nature of each species, understood in terms of its basic needs and need-fulfilling capacities.

The donkey-relative conception of refuse as more choiceworthy than gold is rooted in its highly efficient digestive system, which enables it to meet nutritional needs through scraps and vegetable roughage that a human cannot digest.^[20] It has little use, however, for gold. Humans, in the mind of an Archaic Greek audience, are naturally suited to a life in which we meet our needs not only directly, but also indirectly, through social institutions such as trade.^[21] Gold might appear to be choiceworthy to a creature with such socially-mediated needs; the metal’s homogeneity, its persistence and integrity through time, and its relative scarcity in the natural environment, allow it to undergird a shared value metric.^[22]

The pig-relative conception of mud as more pleasant than pure water is similarly rooted in its natural constitution and needs – more specifically, its poorly-functioning sweat glands. In an area of the world like Western Turkey, where the average daily highs in the summer tend to be in the mid-thirties, the pig’s near-inability to sweat could seriously impair its wellbeing and even its survival prospects. Since the moisture in mud takes longer to evaporate than pure water, access to mud baths is necessary to keeping pigs cool through the hot summer months. By contrast, human flourishing – even, and especially in the summer months – is aided by the baths in clean water that enable us to remove mud, along with any insects and pathogens it might contain. For us, then, pure water is more pleasant than mud.

Finally, the fish-relative conception of seawater as purer than freshwater is also rooted in observations concerning its apparent constitution. This is not merely because saltwater fish cannot survive in freshwater, but also because they tend to be both larger and more vigorous than freshwater fish. This is juxtaposed against the human-relative conception of freshwater as purer than seawater, since only the former is drinkable and life-sustaining for us.

B61, B9 and B13, then, draw upon the audience’s pragmatic beliefs in ways that motivate the unity reading. In all three juxtapositions, the creature-relative opposites are recognizably aligned with each creatures’ own natures – that is, their conditions of survival or flourishing. This tends to suggest that the opposites are not merely an expression of each creature’s idiosyncratic nature, but are a real feature of the object. In other words, all three juxtapositions draw upon the audience’s pragmatic assumptions to motivate the move from qualified opposites (e.g., ‘seawater is pure for a fish’) to unqualified opposites (e.g., ‘seawater is both pure and polluted’).

We might argue that once the audience has added the qualifiers, this move has been rendered unnecessary. They do not need to arrive at the unqualified opposites in order to see to how the two opposites are both really attributable to the object. This is the view of McCabe, upon whose twin principles, ‘unity of opposites’ and ‘opposition of unity,’ I draw in my analysis of the implicit semantic structure of the unity reading and values reading, respectively. On McCabe’s account, an initially paradoxical statement attributes the unqualified opposites to a subject matter (here, “sea is pure and polluted”). This is followed by a resolution of the paradox through the addition of qualifiers (here, “for a fish …” and “for a human …”) that show how the opposites can be attributed to the object without violating common sense and undermining meaningful speech. The first move corresponds to a principle that she calls ‘unity of opposites,’ which emphasizes the connection of the opposites. The second move corresponds to a principle she calls the ‘opposition of unity,’ which emphasizes the difference between the opposites.^[23] On McCabe’s view, B61 offers a kind of hermeneutic for this process by expressing both the unqualified opposites and the qualifiers together.

I agree with McCabe that some such process of aligning the paradoxical claim with common sense must be at work in the audience’s initial encounter with illustrations of the unity of opposites such as B57: “The way of cloth-carders is straight and crooked”’ (γνάφων ὁδὸς εὐθεῖα καὶ σκολιή). Heraclitus can only motivate the unity of opposites, and can only keep the audience engaged in reasoning out his sayings, if he speaks in a way that is both comprehensible and makes some appeal to their intuitions. This suggests that an approach that yields “sense over nonsense,” as McCabe (2015, 45) puts it, must be foundational. In the case of B57, the sense depends upon implicit qualifications that bring in different facets of carding (making straight by aligning the fibers; making crooked by blending the fibers together). Yet this sort of instance of the unity of opposites focuses on the object itself; it does not draw attention to the possibility that the attributions of the opposites might be relative to some unreliable ‘standard.’ By contrast, in qualifying the opposites via creature-relative needs and preferences, the human-animal juxtapositions introduce worries about the ‘standard’ with which these attributions are associated. In doing so, they also ask the audience to consider whether the object might be grasped in a manner that would not be susceptible to the same sorts of worries.

This means, initially, that the audience must be able to use the fact that baths in water are more pleasant for humans than baths in mire, along with the fact that the converse is true for a pig, to refer back to the creature-independent nature of water. A twenty-first century interpreter might do this by talking about properties of mud and water that underlie the different effects that each have on humans and on pigs (e.g., speed of evaporation). A later Classical interpreter might explain the compresence of the ‘pleasant’ and ‘unpleasant’ in mud and water by appealing to Aristotelean-style potentialities or powers which are latent within each, and which are only actualized through contact with some particular constitution.^[24] Either analysis yields a direct link between, on the one hand, the idiosyncratic creaturely ‘natures’ of pigs and humans, and the creature-relative conceptions of the object to which they give rise, and, on the other hand, the nature or properties of the object (i.e., water). The relative or qualified claims about the opposite effects of the object can, on these analyses, be referred back to the independent or unqualified properties of the object, which manifest in different ways under different conditions.

This solution relies, however, on a more sophisticated conceptual apparatus than may have been available to a late Archaic audience. The conception of potentialities builds upon a pre-existing conceptual framework that developed over the Classical period. In Heraclitus’ time, by contrast, and as widely noted by commentators, even the distinction between property and substance was still conceptually blurry.^[25] Moreover, in the generations after Heraclitus, we encounter philosophers who seem to aim, much more directly than he does, to explain how and in what sense objects have the qualities that they have; the contours of the problems addressed by them might have looked much different if the concept of latent potentialities had been available to them. For instance, Democritus has a plausible claim to being the first to formulate a clear conceptual distinction between observer-qualified and observer-independent properties – yet he can only do so by relegating the former to unreality.^[26]

A second, related set of concepts that might have shaped the contours of Heraclitus’ ‘argument’ are those associated with the various psychic faculties and physiological processes. For instance, like the sceptic Sextus Empiricus, Heraclitus contrasts the needs and preferences of humans and nonhuman animals, as manifested through choice and avoidance. Sextus, writing at the cusp of late antiquity, can draw upon (without necessarily endorsing) a range of Platonic, Aristotelean, Stoic, Epicurean and Hippocratic models for differentiating the faculties and explaining the processes that mediate between a creature, an object, and the creature’s responses to that object. In Heraclitus’ time, this kind of modelling was still in its infancy; for instance, it remains controversial whether and how Heraclitus distinguished between thought and perception.^[27] This, especially when combined with the generally deterministic Archaic picture of mortal natures, may be an important factor in the different inferences that the two thinkers motivate through an appeal to a creature’s observed patterns of choice and avoidance (or its species-relative needs and preferences).^[28] In particular, Heraclitus does not distinguish between such instinctive behaviours and value judgements. In fact, as I show in section 3, he attacks his audience’s values and their preferences and needs simultaneously by tracing them back to a creature-relative nature.^[29]

If some of the conceptual models and tools of later philosophy were unavailable for Heraclitus’ use, he nonetheless had at his disposal some powerful cultural assumptions and starting-points that might serve as motivators in directing his audience’s reflections. In the context of the human-animal juxtapositions, the audience moves from their pragmatic assumptions and beliefs concerning the creature-relative qualities of an object to an attempt to grasp an unqualified claim about that object. In making this attempt, as I suggest in the next section, they are likely to find the unity reading collapsing into the values reading – especially given the contribution of a second set of ‘commonsense’ assumptions rooted in a shared system of symbolic associations.

3 Symbolic Assumptions and the Species-Relativity of Axiological Opposites: From the Unity Reading to the Values Reading

If the unity reading is supported by pragmatic beliefs that contribute to the audience’s conceptions of the natures of creatures and objects, it is at odds with a symbolic background that also plays a significant role in structuring those conceptions. The pragmatic beliefs allow for a recognition of the alignment of the two creature-relative conceptions, both of which concern the creature’s basic conditions of survival and flourishing; on this basis, the audience can affirm both of the two qualified opposites. On the other hand, as I argue in this section, the symbolic background emphasizes a marked asymmetry between the animal-relative opposites, regarded as a product of an idiosyncratic, reality-deficient nature, and the human-relative opposites, regarded as simply corresponding to how things are. The assumptions derived from the symbolic background provide an initial standard for non-idiosyncratic assessments of axiological qualities such as pleasant and unpleasant, choiceworthy and worthless. They undergird a pre-reflective view upon which these and other basic quality attributions are simply taken as part of the creature-independent (non-relative) nature of the object just so long as they are not conditioned by a recognizably unreliable or idiosyncratic subject.

As I argue in this section, the attempt to grasp the unity reading challenges this approach to axiological qualities without providing the audience with a satisfactory alternative. The qualified opposites can be grasped as a pair of creature-relative truths, yet this pair ultimately comes apart; it cannot reveal the nature of the object. The careful alignments that motivate the unity reading do, however, reveal the equivalence of the two creature-relative conceptions of the object – with the result that the unity reading structures and gives way to the values reading. Both the animal and human conceptions seem to be idiosyncratic to the natures of the creatures in question, since both are similarly rooted in the relevant creatures’ physiological conditions of survival and flourishing. Both seem to be merely relative, in the sense that the associated axiological opposites are expressive simply of the creature-relative natures upon which they depend.

As outlined in section 2, the unity reading asks the audience to grasp the compresence of the unqualified opposites ‘choiceworthy’ and ‘worthless’ in gold (as well as refuse) and to grasp the unqualified opposites ‘pleasant’ and ‘unpleasant’ in pure water (as well as mud). There is reason to think that Heraclitus’ audience would find the idea of unqualified axiological opposites less peculiar than we do – especially given that the objects used in these juxtapositions occupy extreme poles on a traditional Archaic spectrum of value. In other words, these are objects whose axiological qualities would have seemed uncontroversial – even to the point of functioning as a touchstone for thinking about value and disvalue.

To the Archaic Greek, gold and clean water were not only things that humans happened to need or prefer; they were objects that possessed the quality of value to such a high degree that they could function as axiological emblems; value was a constitutive element within their natures.^[30] A conception of gold as intrinsically or independently valuable was perhaps most visible in the belief that it was valued not only by humans, but also by the gods. Homeric deities regularly appear surrounded by golden garments, furnishings, and implements, while gold and gold-plated objects were traditional votive offerings to the gods at temples and shrines.^[31] Gold also figures strongly in intersections and transactions between the human and divine realms more generally; for instance, when a golden sheep, ram or vine appears in myth, the transposition of the precious metal onto the ordinary symbol of agricultural wealth confers on the latter an affirmation of the divinely-conferred nature of that wealth and the power (and fate) that goes with it.^[32] Pure or potable water was also a commonplace emblem of value, not only because of its role in sustaining life and its relative scarcity, but because it manifests the characteristic of purity to such a high degree that it also confers it upon other objects (i.e., through washing). As Gerber notes, Pindar’s observation that “Water is best” (Ἄριστον μὲν ὕδωρ) at Olympian 1.1 expressed what was likely a self-evident example of something that was valuable “in an absolute sense” (Gerber 1982, 7–8).

At the other end of the evaluative spectrum, sweepings, or σύρματα – used, as in B9, to refer to what is cast off (including chaff and other inedible roughage) – are the dregs left over when whatever is of value has been removed. In B13, the onomatopoeic term βόρβορος (mud, mire, filth) similarly draws attention to the disvalue of the polluting and polluted character of the stuff that the pig enjoys instead of pure water.^[33]

The symbolically-structured conception of gold, water, refuse and mire, then, fits well with a pre-reflective view of the (dis)value that belongs to each, unqualified by any creature-relative needs and preferences. This pre-reflective view need not, initially, be challenged by the fact that the putatively creature-independent axiological qualities significantly overlap with human-relative considerations of usefulness, desirability or repugnance. On the pre-reflective view, the human-relative preferences are not restricted to humans; after all, humans and gods share the same preferences. They are thus not really human-relative – at least not in the same sense as the pig’s preference for mire is relative to it. More precisely, human preferences in these cases simply do not raise the issue of creature-relativity to begin with, since such preferences are presumed as a reliable standard of the (dis)value of such familiar objects. By contrast, the animal-relative needs and preferences can be easily dismissed as idiosyncratic to the natures of each – especially given the fact that traditional representations of donkeys, pigs and fish emphasized the kinds of faulty natures ill-suited to accurate evaluative discernment.

The donkey, for instance, was traditionally depicted as ignoble and ignorant, ruled by appetites that left it insensible of all other considerations beyond their immediate satisfaction.^[34] These same appetites were also built into the donkey’s proverbial lack of discrimination – so that its ability to benefit from garbage that humans cannot even digest could easily be dismissed as a sign of a faulty, indiscriminate nature, as could its inability to recognize or benefit from an ‘inherently valuable’ object like gold.

Similar observations might be made regarding the pig-relative conception of mire and pure water in B13 and the fish-relative conception of seawater in B61. The poetic tradition associated pigs, like donkeys, with natural ignorance, lack of discrimination, and failure to appreciate higher things. Like donkeys, they have ‘indiscriminate’ eating habits – they can be nourished on scraps, and even human waste.^[35] Their proverbial lack of culture and stupidity formed the basis of the pejorative expression ‘Boeotian sow’ (Βοιωτία ὗς) in Archaic and Classical poetry.^[36] Fish, on the other hand, were associated with ignorance on the basis of their silence, since the early Greeks treated speech and thought as one and the same – regarding the mute, for instance, as cognitively deficient.^[37] In the handful of appearances of fish in the Homeric epics, those that show them in their own element (rather than flopping helplessly on land) tend to emphasize appetites that are viciously indiscriminate.^[38]

The symbolic background that helps structure the audience’s conception of axiological qualities, then, lends itself to the view that opposites such as ‘choiceworthy,’ ‘worthless’, ‘pleasant’ or ‘unpleasant’ are intrinsic to, and partially constitutive of, the natures of gold, water, refuse and mud. This background provided an immediate source of truisms about what an object was really like, axiologically speaking, rather than merely how it functioned relative to the needs or preferences of a particular creaturely nature. On the symbolically structured picture of the axiological qualities, the human-relative opposite is constitutive of the independent nature of the object. The animal-relative opposite, on the other hand, is merely relative. It attests, not to the characteristics of the object as such, but to the limitations of a creaturely nature that exhibits an absence of evaluative discernment on the level of basic physiology as well as its cognitive and behavioral manifestations.

However, the unity reading is incompatible with this one-sided view. Again, as argued in section 2, the animal-human juxtapositions are structured in such a way as to invite the unity reading through careful alignments and parallelisms. The recognition of this structure in the human-animal juxtapositions places a demand upon the audience that they seek the compresence of the opposites in gold, refuse, water and mud. In order to do this, they must find a way to move from the qualified opposites (which merely express the natures of the two creatures) to the unqualified opposites (which express a truth about the object). It is only through this process that they grasp the real axiological nature of the object. As noted in the introduction, the epistemic significance of this demand is underscored by the fact that Heraclitus presents the failure to grasp the unity of opposites as a demonstration of ignorance about the apparently familiar object.

Yet, as argued in section 2, it is difficult to see how, using the conceptual distinctions available to them, the audience can move from creature-dependent observations to the creature-independent object. That is, they seem to lack the kind of intermediate concept or model that would allow them to grasp the equal, unqualified opposites of the unity reading as compresent in the object. Instead, they are left with the equivalence of the two separate creature-relative opposites. For instance, rather than grasping the equal presence of the opposites ‘unpleasant’ and ‘pleasant’ in mud, they recognize the equivalent discernment manifested by humans and by pigs in responding as though mud was unpleasant or pleasant, respectively. Given that the audience has not managed to eliminate the creature-relative qualifiers, the implications of this alignment are hard to avoid. Human evaluative discernment now seems to be on the same footing as that of creatures whose responses and preferences the audience is accustomed to treating as a mere by-product of a base, ignorant nature. This is reinforced by the fact that the parallelism lays bare the idiosyncratic – that is, species-dependent – ground in which human and animal values alike are rooted.

By emphasizing the equivalence of the animal-relative and human-relative conceptions of the value of water, mud, gold and refuse, the failed unity reading highlights how both arise in the same way from each creature’s nature – that is, its physiological conditions of survival and wellbeing. The human-animal juxtapositions present the evaluative discernment displayed by humans, pigs and donkeys as similarly driven by idiosyncrasies that belong to their species-nature. To put it another way, the juxtapositions emphasize value ascriptions that are bound up with their limits qua (some particular kind of) mortal creature – a creature that will sicken and die if unable to provide itself with adequate nourishment (B9) or unable to combat the effects of extreme heat (B13). This creates a clear and tightly aligned symmetry between the evaluative discernment of humans and that of nonhuman animals. As such, it encourages a suspicion that neither human nor animal-relative value ascriptions are an expression of the axiological nature of that object; each is merely an expression of each creature’s nature.

At this point, the unity reading seems to have collapsed into the values reading. Nonetheless, an audience that has accepted at least some of Heraclitus’ lessons about the unity of opposites might not rest easily here. The implications of the values reading, with its emphasis upon human ignorance of the objects’ true axiological natures, maps too directly onto the notion that the failure to grasp the unity of opposites with respect to some familiar object constitutes ignorance of that object. If the audience’s ignorance lies only in this failure, then there seems to be some hope for them. Perhaps there is, in fact, some degree of insight into the object which can be found in both the human- and animal-relative conceptions – provided, that is, that the two can be brought together through the unity reading. This approach, which I will call the ‘partiality interpretation,’ holds that the audience’s lack of evaluative discernment equals that of nonhuman animals only so long as they continue to cognize the objects’ axiological natures from an anthropocentric vantage of the kind exemplified by the symbolically-structured conception of qualities. It suggests, paradoxically, that there is a sense in which the evaluative discernment that belongs to human nature is superior to that which belongs to animal nature – but only insofar as it allows the audience to engage with animal-relative values in such a way as to overcome human habits of thought and valuation.^[39] As long as they remain tethered by these habits, they will be unable to see the axiological opposites as equally salient or as equally present in a given object.

The problem, at this stage, is that B9 and B13 have given the audience a picture of value ascriptions framed entirely by the use that each object serves for each creature given its species-nature. The axiological opposites themselves are visible only through and in relation to each creature’s idiosyncratic constitution – and, more specifically, through the ways in which a given object is useful or useless in fulfilling that creature’s natural needs and preferences. In other words, Heraclitus has presented the axiological opposites in such a way that they offer nothing to hold onto beyond these natural needs and preferences. In this context, it looks as though the audience has little hope of grasping value and disvalue in their equality as compresent opposites in an object; the human conception of these opposites will always skew heavily towards their own species-relative needs and preferences.

This need not be fatal to the unity reading, so long as the audience can find some other way of understanding these opposites that does not depend upon the object’s merely creature-relative value. The species-relative axiological opposites may ultimately map onto a pair of opposites that is not merely constituted by creaturely needs and preferences. If this is so, then the oppositions in B9 and B13 would express only an intermediate state, and their unachieved unity would point the audience away from the kind of thinking immediately constrained by human nature and towards a true unity of opposites. If this is the case, the strongest evidence – and best guide for how the process might work – would be found in B61.^[40] In this fragment, we seem to encounter just such a mapping between, on the one hand, opposites immediately rooted in species-dependent needs and preferences, and, on the other hand, opposites apparently independent of creaturely natures. The problem, as I argue in the next section, is that the relation between the species-relative opposites and species-independent opposites in B61 does not lead towards a grasp of the unity of opposites. That is, it fails to lead the audience out of their human-skewed approach to the species-relative opposites and into the impartial unity of species-independent opposites. Rather, it shows how the species-independent opposites are not only tied to, but ultimately reduce to, the species-relative opposites, which are exhaustively constitutive of their meaning.

4 From Species-Relative Opposites to Species-Independent Unities?

At this stage, the audience might have some motivation for following any new approach that might allow them to grasp the unity of opposites in the human-animal juxtapositions. If they can do this, then they may be able to overcome the damning assessment of human epistemic capacities with which the version of the values reading sketched in section 3 leaves them. In this respect, B61 seems to offer some hope. This fragment, again, runs as follows: “The sea is the purest and most polluted water: drinkable and life-preserving for fish, but undrinkable and deadly for humans.”

The (ostensibly) species-independent opposites ‘pure’ and ‘polluted’ in this fragment map onto a pair of species-relative qualifiers. The latter are akin to the opposites of B9 and B13: they are aligned through their parallelism and opposition, and are expressed in terms of natural, species-relative needs (in this case, ‘drinkable,’ ‘undrinkable,’ ‘life-preserving,’ ‘deadly’). As McCabe observes, the key question for dealing with these sorts of qualifiers is the direction of inference.^[41] Do the species-relative opposites ‘drinkable’ and ‘undrinkable,’ ‘life-preserving’ and ‘deadly’ guide the audience away from the (merely) creature-relative and towards the true nature of seawater by their connection to ‘pure’ and ‘polluted’ – or do ‘pure’ and ‘polluted’ turn out to be reducible to the partiality of ‘drinkable,’ ‘undrinkable,’ ‘life-preserving,’ and ‘deadly’?

In this section, I argue that, rather than showing the audience how to free the species-relative opposites from the partiality of creaturely needs and preferences, B61 presents the ostensibly species-independent opposites – and even the water in which they are supposed to inhere – as suffering the same fate as the species-relative opposites. The qualities that might seem to inhere in water on a species-independent basis, and even the category ‘water,’ reduce to distinctions and quality ascriptions that are relative to creaturely needs and preferences. This has two aspects. First, the basic character of the ‘species-independent’ opposites ascribed to water breaks down to the overtly species-relative opposites. Second, the category ‘water’ is found, via the attempt to understand it in terms of those ostensibly species-independent opposites, to be itself defined in relation to narrow, creaturely needs and values.

In order to support the partiality interpretation, we would need to find that the inference in B61 moved in the other direction. The concepts of purity and pollution, while graspable via the species-relative opposites, would also have a significance that went beyond the merely species-relative. More specifically, ‘pure’ and ‘polluted’ would have to be grasped as equal, compresent opposites that belong to and express something about the nature of seawater, rather than just the natures of the creatures who need it or avoid it. I argue, however, that this is not the direction of inference in B61. On the contrary, the reduction of the species-independent opposites to the species-relative opposites is further reinforced by the reduction of water to the opposites ‘pure’ and ‘polluted’ – or, more precisely, to the needs and preferences that are inextricably bound up with defining an object such as water, not only as ‘pure’ and ‘polluted,’ but even, simply, as ‘water.’

B61 begins by unifying the opposites ‘pure’ and ‘polluted’ via their compresence in seawater. This pair of ostensibly species-independent opposites has been aligned with two other pairs of opposites, whose relativity to a given set of creaturely needs and preferences is flagged by the fact that they are enclosed within clauses qualified by the datives “for fish” (ἰχθύσι) and “for humans” (ἀνθρώποις), respectively. As argued in sections 2 and 3, a juxtaposition of this sort eludes the unity reading, failing to speak to the nature of the object, if the opposites are simply a function of idiosyncratic creaturely natures. In B61, this means that in order for the partiality interpretation to succeed, the audience must find a way to understand the species-independent opposites of the main clause (‘pure,’ ‘polluted’) that does not simply reduce them to the species-relative content of the qualifying clauses (‘life-preserving,’ ‘deadly’; ‘drinkable,’ ‘undrinkable’). Specifically, the species-relative opposites must lead away from their partial starting point, and towards a broader vantage not defined simply by creaturely needs and values. If this process is found to be at work in B61, it could indicate a similar process is implicit in B9 and B13.

It seems possible that such a movement might be identified via the link between the nature of water and the use that it serves for particular creaturely needs and preferences. This is the approach taken by Kahn in his ‘partiality interpretation’ of the human-animal juxtapositions.

On his interpretation, the use that water serves for creatures with specific natures reveals the real function and nature of water as such. More specifically, the species-relative qualities that allow water to serve the needs of fish and humans (potability, life-preserving character) also manifest the species-independent qualities (pure, polluted) that, in turn, manifest the true nature of water itself.^[42]

The first step – the link between the species-relative qualities and the species-independent ones – is well-supported. Heraclitus’ use of the term καθαρός seems to amount to ‘without admixture’ (LSJ, s. v. I.2), which can plausibly be glossed as ‘sound, whole (and therefore usable for its purpose)’ (LSJ, s. v. I.6). In this case, the purity or pollution of water is expressed through its ability or inability to do what water is supposed to do. On this basis, one might call seawater either ‘pure’ considered relative to its usefulness in fulfilling the needs of fish, or ‘polluted’ considered relative to its failure to fulfill the needs of humans.

On its own, however, this first step merely reduces the ostensibly species-independent opposites, ‘pure’ and ‘polluted,’ to species-relative opposites. In order to show that the movement goes in the opposite direction – that the species-relative opposites find a non-relative basis in species-independent opposites – the juxtaposition’s manner of aligning purity and pollution must express something about the true nature of water. More precisely, it should do so in such a way that the audience can actually grasp what is expressed, as they could not (directly) grasp the opposites ‘choiceworthy’ and ‘worthless’ as an expression of the nature of refuse.

Kahn thinks that this is, in fact, the case. The question of whether water is pure or polluted, whether it can function properly as water for a given creature, is bound up with the question of whether it is properly water, independent of creaturely needs. He writes:

Thus we are both right and wrong to perceive the sea water as water […] For it is water in physical or cosmic terms […] But it cannot function as water for the vital needs of men. This is a limitation of human nature, however, not a defect of the sea. Its virtue as water is manifest in the life of fishes. (Kahn 1979, 187)

Yet it is not clear how far this reading can support the view that the species-relative uses of water reveal its species-independent nature. It looks as though, on Kahn’s interpretation, the sea really is water in the fullest sense. To say that humans would be (partially) “wrong to perceive seawater as water” is merely to express a limitation within our own idiosyncratic natures; on this analysis, the conception of seawater as ‘polluted’ collapses entirely back into species-relative needs and values. Fish nature, by contrast, does not exhibit the same limitations with respect to seawater; fish respond to the non-species relative (“physical or cosmic”) nature of the sea-as-water correctly. The unity reading breaks down here, as does the promise that humans might reliably discern the true nature of water by using our natural needs and values as starting-points.

One need not accept Kahn’s cosmological interpretation of seawater in B61 – or the more extreme version of the values reading that seems to follow from it – to recognize the plausibility of the use-centered interpretation of ‘pure’ and ‘polluted’ in this fragment. However, without a viable bridge between the true and species-independent nature of water, on the one hand, and the use that water serves given a specific set of creaturely needs and preferences, on the other hand – without, in other words, a transition from the values reading to the unity reading – the focus on function merely underscores the reduction of the species-independent qualities to the species-relative ones. Kahn is surely correct in his assumption that, for Heraclitus’ audience, seawater was recognizable as a form of water. B61 assumes this as a starting-point. Its initial challenges do not concern the question of whether – and in what sense – the sea is water. These questions come to the fore, however, after ‘pure’ and ‘polluted’ have been reduced to the needs that water fulfills for humans given our idiosyncratic species-natures.

Pure water is, again, water in the fullest sense; it is ‘without admixture’ (LSJ, s. v. I.2) – that is, uncontaminated by what is not natural to it, by what is not water. However, by using the qualifying phrases ‘for fish’ and ‘for humans’ to unpack the notions of ‘pure water’ and ‘polluted water,’ B61 emphasizes that what counts as an additional, polluting element depends in fundamental ways on what we need or value something for. As the values reading has shown, there is no reason to think that human needs and preferences are well-matched to the true character of things. In tying the purity of water to the function that it serves for a particular creature, then, Heraclitus opens the door to fundamental questions about the nature of water and of the opposites more generally.

The notion of an object’s purity plays a role in the deeper layer of epistemic puzzles not only in B61, but also in B9 and B13. In the next section, I look at how these challenges manifest in the latter two fragments. I argue that the challenge to the human judgements created by the human-animal juxtapositions is broader in scope than the values reading would imply. Mediated (initially) by the unity reading, with its demand to unify the opposites, the animal-human juxtapositions do not merely undermine the judgements obviously tied to human nature and anthropocentric values. The juxtapositions challenge the very process of distinction-making, showing how the audience’s physiologically determined needs and preferences guide the process by which they distinguish water from non-water, and by which they pick out discrete objects more generally.

5 Beyond the Values Reading: The Species-Relativity of Distinctions

At this point, the audience has been forced to give up the unity reading of the human-animal juxtapositions. Even if it is true that the comprehension of the nature of the object depends on a grasp of the opposites unified within it, it is far from clear how they are to go about grasping those opposites. At first, the problem seemed to be the fact that their nature as humans prevented them from fully embracing the unity of axiological opposites such as ‘choiceworthy’ and ‘worthless’ (with respect to refuse or gold) or ‘pleasant’ and ‘unpleasant’ (with respect to pure water or mire). Now, however, it appears that such species-relative opposites fail to map onto the species-independent nature of things; moreover, they seem to determine apparently species-independent opposites, which, in turn determine how creatures distinguish objects such as water, refuse and gold. As a result, it is not only value judgements that are called into question, but more basic differentiations. The latter, too, seem to be idiosyncratic to species natures.

We can see this more clearly by examining the role of the concepts ‘pure’ and ‘polluted’ in B9 and B13. Both in B13 (where the term καθαρός appears explicitly) and in B9 (where it does not), Heraclitus’ choice of objects draws attention to notions of purity and pollution in relation to distinction-making. Like B61, these fragments involve objects whose species-relativity turns out to be constitutive of the purity or impurity of some specific kind of object. As in B61, this species-relative conception of the object plays a crucial role in distinguishing it from other objects.

As noted above, a crucial meaning of καθαρός in these fragments is ‘clear of admixture’ (LSJ II, s. v.). Again, this would seem to be the primary sense in which the notion of ‘purity’ informs distinctions between things, since the description of a thing as pure or impure reveals what we think a thing is: in describing an object as impure, we implicitly invoke a conception of how that object may be (or fail to be) fully or exclusively itself. It is telling, then, that of the four materials listed in LSJ as paradigmatic cases of what is properly ‘pure,’ three are represented among the four different objects deployed in Heraclitus’ animal-human juxtapositions (water, winnowed grain, metals). What is particularly striking about these examples is that the issue of their purity foregrounds not only natural needs and preferences, but also the modes of distinction-making that arise from them.

The most obviously paradigmatic case, that of pure water, appears in both B61 and B13. The purer water is, the more it is properly water; the more polluted water is, the less it is properly water. Yet as we saw in section 4, the purity or impurity of water is also bound up with physiological limits and capacities that differ from species to species. In relation to these, as we saw in section 3, the distinctions made by humans are not to be privileged over those made by nonhuman animals; human nature is no more closely matched than that of nonhuman animals to the true character of things. In B61, pure water is water that is ‘life-preserving’ and ‘drinkable.’ The impurity of seawater for humans lies in the fact that we cannot drink it – for us, it cannot serve the function of water, since it is polluted through the addition of salt. Yet for saltwater fish, most of which cannot survive in freshwater, our ‘pure water’ would be water from which salt had been subtracted, and which, in the subtraction, had become polluted and death-bringing. In B13, pure water is water in which we can wash ourselves in a way that decreases susceptibility to heat exhaustion and makes us less attractive to disease-carrying insects such as mosquitoes. For us, the impurity of mire, or water mixed with earth, lies in the fact that we cannot wash ourselves in it and gain relief from the itching and heat of summer temperatures and insects. Yet for a pig, the purity of mire lies in precisely the fact that it is a better source of relief from heat and parasites than water alone. In B13, then, as in B61, a set of immediately species-relative opposites (‘pleasant’ and ‘unpleasant’) turn out to be constitutive of the ostensibly species-independent opposites, ‘pure’ and (implicitly) ‘polluted.’ In B13, as in B61, the application of this second pair of opposites to specific kinds of objects shows how these objects are distinguished in accordance with the creaturely needs and values expressed in the first, species-relative set of opposites.

In fact, in relation to this second layer of challenges, B13 seems to provoke even deeper worries about our ability to distinguish between water and non-water. While B61 raised questions about the true nature of water by problematizing the distinction between pure and impure water, B13 raises questions about what even counts as water. Humans do not recognize mire as water – unlike seawater, water mixed with earth has been polluted to the point that it is no longer the same kind of stuff as the water we drink and in which we bathe. The fact that pigs treat mire as a form of water, where humans do not, is a result of each species’ physiologically-grounded needs and preferences, the conception of purity that these needs and preferences yield within a given context, and the categories and distinctions that this conception, in turn, generates and serves.

This idea that the distinctions humans make among objects are an expression of their species-dependent needs and preferences is further supported by B9. This fragment, again, tells us that donkeys prefer refuse to gold. The audience’s pre-reflective assumptions would simply have taken their own preferences, as against those of the purportedly ignorant, ignoble donkey, as indicative of the objects’ true axiological qualities. Further reflection on the juxtaposition, as we saw, not only challenges this easy assumption about the priority of the human-relative over the donkey-relative conception, but suggests that both conceptions might turn out to be hopelessly idiosyncratic. Revisited in light of B61’s challenges, B9 offers a direct illustration of how our needs and preferences not only guide our attribution of such qualities as ‘choiceworthy’ and ‘worthless’ to a given object, but also shape our modes of distinction-making at a more fundamental level. More specifically, it illustrates how evidently species-relative ascriptions map onto conceptions of purity and pollution, and how the latter, in turn, are bound up with whether we even recognize something as a distinct object.

Of the two kinds of objects featured in B9, sweepings or σύρματα are a paradigmatic instance of a polluted medium (like mire); gold is a paradigmatic instance of a pure one (like clean water). In fact, there is a sense in which the term ‘sweepings’ does not name a specific kind of thing at all, but only an indefinite mass – a mixture of household and agricultural refuse (in particular, the straw, husks and pests removed by the threshing and winnowing required for the production of a pure, valuable and nourishing grain). These varied stuffs are of so little use in fulfilling human needs, and possess so little perceptual salience for us that, once they are heaped up together as sweepings, we are often unable to distinguish the component parts. The identification of σύρματα as a particular kind of object arises from the fact that, being made up of the discarded parts of other objects, it gathers together all the impurities and worthless elements that detract from the value of the latter. In other words, its distinguishing feature, for humans, is the fact that it pollutes other things. This places sweepings and gold on opposite poles not only with respect to such qualities as disvalue/value (or worthlessness/choiceworthiness) and pollution/purity, but also with respect to issues of differentiation. Gold, like clean water, is, in a sense, only fully itself when it is pure – its familiar qualities and its value are attained through a process of differentiation and distillation in which its ores are extracted from the earth (or the water) and distilled out through ore-washing and smelting. Moreover, in contrast to sweepings or refuse, gold draws the human eye. This is, in part, a result of the value of gold in fulfilling needs in human marketplaces and communities – though it is also bound up with human responsiveness to colors, to shiny, light-refracting surfaces, and to visual stimuli in general.^[43]

Thus, for the audience, it is neither possible nor worthwhile to distinguish between the various elements contained within a heap of sweepings; it is, however, both possible and important to distinguish gold from other objects within the landscape, and from other metals. For donkeys, the converse is true. Their hardy digestive systems make it possible for them to nourish themselves on the chaff and roughage that are, for humans, fit only for compost; the donkey’s highly developed senses of smell and taste give it the capacity to distinguish between different elements of what is for us an undifferentiated heap. Donkeys, on the other hand, lack the strong responsiveness to visual stimuli, along with the communal structures of currency and trade, that allow gold to draw the audience’s attention. In fact, for the donkey, gold likely belongs to the same sort of category of undifferentiated heaps that humans mark as σύρματα. To a human observer, at least, donkeys might seem not to differentiate between gold and other parts of their environment that are neither useable for nor an impediment to fulfilling their needs.

In both B13 and B9, then, we find the axiological opposites pleasant/unpleasant and choiceworthy/worthless serving much the same function as the species-relative opposites life-preserving/deadly and drinkable/undrinkable in B61. In all three fragments, the more immediately species-relative opposites condition the meaning of the ostensibly species-independent opposites, pure/polluted. The latter pair of opposites, in turn, conditions how humans conceive of a given kind of object, and how (and whether) they distinguish it from other things. B61 is constructed in such a way that the different kinds of challenges can be seen in relation to one another: the less obviously species-relative opposites are attributed to seawater directly, while the opposites immediately expressive of needs and preferences (drinkable/undrinkable; life-preserving/deadly) are set off in qualifying clauses that specify the creature (fish or human) that possesses those needs and preferences, relative to which seawater is pure or polluted.^[44] As such, this fragment also offers a clear illustration of the relationship between the conception of water as ‘pure,’ the conception of water as (really) water, and the use that water serves for some particular creature with a particular nature. B9 and B13, on the other hand, intensify and deepen the epistemic challenges raised by B61. First, these fragments show that the problem is not restricted to a single kind of object – the same problems that B61 introduced with respect to water are now seen to exist for mire, gold and sweepings as well. Second, we are no longer restricted to the qualities or natures possessed by some determinate kind of object. In other words, we are not merely, as in B61, concerned with the kind of alterations that would make an object less pure, or less itself. In B13, the question is whether a given medium even qualifies as water. That is, the issue is not whether ‘water plus salt’ is pure water or ‘water plus non-water,’ but whether ‘water plus earth’ is water or non-water; we are concerned, in this case, not with how and where an object becomes less itself, but where it ceases to be one kind of object and becomes another kind of object entirely. In B9, the problem is perhaps given its starkest expression. In this fragment, the needs and preferences that belong to human nature – as well as the need-fulfilling sensory capacities with which they are bound up – seem to determine whether, in a given case, we even make the kinds of distinctions involved in picking out and identifying objects as anything at all, rather than treating them as an undifferentiated mass.

6 Future Questions: Human Language and Heraclitean ‘Scepticism’

The problem, then, turns out to be deeper than the idiosyncratic character of human values. Idiosyncratic, human-relative assessments concerning axiological opposites (section 3) are also the basis for ascriptions of apparently non-axiological, species-independent opposites (section 4) and for the differentiation of objects (section 5). In other words, the differentiations expressed in basic category distinctions cannot be separated from the evaluative distinctions expressed in B9 and B13 in terms of the ‘choiceworthy’ and ‘worthless,’ the ‘pleasant’ and ‘unpleasant.’ These evaluative distinctions, in turn, reduce to such natural ‘volitional distinctions’ as are expressed in the instinct towards choice and avoidance.

As indicated in section 1, this article has followed an interpretive model that emphasizes Heraclitus’ deliberate exploitation of multiple layers of possible meaning. This approach, again, involves an assumption that the audience is meant to reconsider, discard, and amend previous readings – or at least repurpose them under a new aspect. Since this means that many readings will turn out to be provisional, my argument is consistent with the possibility that Heraclitus, ultimately, intends the reading encouraged by the human-animal juxtapositions to be revised in light of other sayings.^[45] It is, for instance, consistent with the view that the deepest layer of challenges functions primarily as a goad to the audience to reconsider cherished assumptions. Equally, however, my argument is consistent with the possibility that the challenge to human distinction-making is meant not only to stand but even to be deepened, problematizing beliefs that constituted the juxtapositions’ starting-points – for instance, the belief in species-natures and in the opposition-based structures through which the audience sought the nature of objects.

The choice between these (and other) options rests, in part, upon how we interpret the interrelation between creature-relativity and language in Heraclitus’ fragments. This invokes a raft of controversies – including those around logos and names – that puts any decisive answer beyond the scope of this article. However, even a brief consideration of these relationships brings into focus some of the questions towards which the juxtapositions seem to lead us – while, at the same time, allowing us to see the scope of the epistemic challenge that the juxtapositions present even if read as merely provisional. In the final part of this article, then, I consider how the problem of the creature-relativity of distinction-making might be impacted by the characteristics of human language.

Language seems to represent a key element of human distinctions that is absent from the distinctions made by nonhuman animals; the human-animal juxtapositions, in emphasizing the parallelisms between human-relative and animal-relative distinction-making, invite questions about how the resultant epistemic worries might be mitigated or reinforced when we take into account the role that language plays in our differentiations.

On the one hand, the range of linguistically-toned assumptions and tools Heraclitus draws upon, challenges and refines in the animal-human juxtapositions alone might suggest that speech is not idiosyncratic; it might suggest that, simply by adhering to the consistent structural principles of language, we can get closer to the kind of speech that reflects things as they really are.

On the other hand, this apparent generality could actually turn out to be a weakness of language. It might be that language simply enshrines the same sorts of idiosyncrasies that humans attribute to nonhuman animals, under the guise of representing a common world.^[46]

Regardless of the answer we choose, the human-animal juxtapositions in B9, B13 and B61 create a methodological paradox. They affirm the idiosyncrasy and creature-relativity of the distinctions and categories upon which language depends. Yet their challenge to distinction-making proceeds via structures associated with (linguistic) distinctions – in particular, the structures of opposition and analogy. It also trades upon the view that these structures are the basis of a knowledge that is consistent and common rather than idiosyncratic.

This methodological paradox is brought forward directly by a fourth human-animal juxtaposition: B4. The authenticity of this fragment is more controversial than that of the other three fragments discussed in this article.^[47] Yet in light of both its similarity to the other three juxtapositions and its specific focus upon language, its analysis may prove helpful in highlighting the contours of the problem.

B4 runs as follows: “If happiness consisted in the pleasures of the body, we would call oxen happy whenever they come across bitter vetch to eat.”^[48] (Si felicitas esset in delectationibus corporis, boves felices diceremus, cum inveniant orobum ad comedendum.)

The fragment, at least as transmitted, aligns two overlapping terms, ‘happiness’ and ‘pleasure,’ while also setting them within an overarching structure of human speech. If we assume that ‘happiness’ (felicitas) here corresponds to something like εὐδαιμονία or ὄλβος and that ‘pleasures’ (delectationibus) corresponds to something like ἡδοναί, then the terms are not ones that are easily separable in Heraclitus’ time. They do, however, diverge in two ways. First, pleasure is a temporary state that is easily attributed to humans, nonhuman animals, and gods. Happiness, by contrast, is more sustained and stable; this means that while humans can be happy, the gods tend to appear as the source and constant possessor of happiness.^[49] Second, it is possible to enjoy even things that are not truly pleasant^[50]; it is these sorts of ‘pleasures’ that belong to livestock. A happy life, by contrast, is one in which one receives and enjoys the fullest portion of those things that are genuinely pleasant. These are not things that are enjoyed by cattle or pigs, which means that neither the pig nor the cow can be happy. On the popular conception, such things are enjoyed only by humans and by the (anthropomorphic) gods; this also means that genuine pleasures, on this conception, are not species-relative as are those of cattle and pigs.

This also means that there is unlikely to be any strong contrast between pleasure and happiness in this fragment; happiness simply consists in more and better pleasures. The most prominent example of this kind of pleasures – that is, the kind that constitute the happiness of both humans and gods – are the sympotic pleasures of drinking and popular poetry. These sensory pleasures are dependent upon human perceptual responses.^[51] They also, even as popularly conceived, seem to be rooted in human-relative needs – and in much the same way as the human preference for gold (B9) is rooted in ‘natural’ communal structures of trade. More specifically, the sympotic pleasures were part of the model of the sort of unified community that supports survival and wellbeing. The point is, perhaps, underscored by the fact that the cattle’s enjoyment stands out as idiosyncratic from a human-relative vantage not only because bitter vetch was both unpleasant to the taste and difficult to digest, but because it was recognized as a famine food; that is, it was the kind of food that an individual, left without the support of the human community and its happiness-constituting pleasures, would fall back upon.

Conche^[52] argues convincingly that the apparent absence of an alternative model of happiness tends to point away from any substantive contrast between happiness and pleasure in B4.^[53] Can this fragment nonetheless be understood as an instance of the unity of opposites? Conche thinks not, though most commentators who treat this fragment as a possible reminiscence of a lost original do tend to read it in this light.^[54] B4 does seem to gesture towards the sort of unity reading offered by B9 and B13; the audience might at least wonder whether ‘pleasant’ and ‘unpleasant’ might both be attributable to bitter vetch. Unlike B9 and B13, however, however, B4 contains no paradigmatic object of (human) value, only one of disvalue. The absence of the former accents the negative and ensures a swift movement towards the values reading. Human happiness – the blessings of prosperity, peace, drinking-parties with friends, communal approbation – is simply reframed as the pleasure of cattle with plenty of fodder for grazing. Both kinds of pleasures are reducible to markers of a long and comfortable life. B4 differs from the other three human-animal juxtapositions in that it deals directly with the linguistic implications that B61, B9 and B13 create through their challenges to distinctionmaking.

What sets B4 apart from the other juxtapositions, as Conche observes, is its emphasis on what “we would say” (diceremus). In this fragment, he argues, the audience’s own speech becomes a vehicle through which human pleasures are refuted via assimilation to the pleasures of nonhuman animals. We give the name of happiness to the former and withhold it from the latter, but the distinction cannot be sustained; on closer inspection, they turn out to be the same sort of pleasures. The distinction that the audience thought was part of how things really are in the world has, once again, turned out to be rooted in their own creature-relative needs and preferences.

Yet the emphasis upon language in B4 does not simply draw attention to how speech follows and reinforces the distinctions mapped out by creature-relative values. It also, by the same token, draws attention to the role that speech can play in challenging those idiosyncratic distinctions. It does this by explicitly drawing attention to the way the human-relative functions not only as the source of one set of evaluative assessments (‘unpleasant’ as applied to bitter vetch; ‘happiness’ as applied to unspecified human pleasures) but also plays a structural role in aligning those terms with the cattle-relative assessments (‘pleasant’). This, again, is the contribution of the word diceremus (“we would say”). As the hinge responsible for both the contrast and the equating of the two sets of terms, diceremus draws attention to what happens not only in this fragment, but in all of the human-animal juxtapositions: it sets two sides, one human-relative, one animal-relative, within the overarching human-relative framework of speech itself.^[55] This reinforces the fact that the comparison between the human-relative and the animal-relative is rooted in those linguistic distinctions that that same comparison throws into doubt. Yet it also draws attention to the role that the deliberate structural relationship among words, both contrast and similarity or identity, plays in allowing that doubt to register in the first place.

In emphasizing the role that language plays in helping the audience to recognize the idiosyncrasy of language, B4 might fit well with the approach of McCabe. For McCabe, Heraclitus’ method operates through a pair of self-referential structural principles (unity of opposites and opposition of unity) that ensure continued dialectical inquiry by generating paradoxes which generate resolutions, which, in turn, open upon higher level paradoxes. In this case, the idiosyncrasy of language and the commonality (or consistent structural principles) through which that idiosyncrasy is recognized might constitute the source of a higher-level paradox. My analysis of the juxtapositions would support this approach to some extent, since it gives consistent linguistic structures a significant part in challenging distinctions that natural needs and preferences might impose automatically. For McCabe, however, the process of inquiry progressively clarifies the dialectical structure that guides it; the more we attend to these self-referential language-based structures, the more we grasp the implicit principles upon which our common discourse depends.^[56] My interpretation of the human-animal juxtapositions, by contrast, has dealt only with the negative virtues of language. Linguistic structures play a key role in undermining the kinds of category differentiations upon which language depends. This means that language is given a crucial role in allowing the audience to recognize the idiosyncrasy of the human-relative; it does not necessarily mean that language can itself become anything other than idiosyncratically human.

The conception of language as possessing, at best, the power to draw our attention to our own epistemic limits might tally with Burnyeat’s interpretation of Heraclitus. On Burnyeat’s account, however, the possible negative virtues of language are even more strictly circumscribed, because the idiosyncrasies that language helps us to recognize also seem to be those that it creates.^[57] The conventional categories upon which language depends,^[58] and even (perhaps) the view of the opposites as in conflict, are idiosyncratic to the way in which humans, as language users, experience the world.^[59] The solution, according to Burnyeat, lies in the model of the ‘god’s eye vantage’ or absolute vantage within which are synthesized the partial viewpoints belonging to different kinds of creatures on such categories as ‘just.’^[60] This is possible because ‘just,’ from the god’s eye vantage, does not imply the opposition between ‘just’ and ‘unjust.’^[61] However, a category such as ‘just’ is itself a reflection of the human perspective. This means that, like all human categories, its meaningfulness is grounded in relations of conceptually interdependent yet mutually exclusive dualities; ‘just’ and ‘unjust’ cannot be cognized either severed from one another or as co-existing without qualification in a single object. By presenting the absolute or divine vantage as one in which the categories would have to function in ways contrary to their basic conditions of meaning, Heraclitus also presents these categories, and the cognition reflected and circumscribed by them, as inescapably human. This means that, while it appears on the surface as the single viewpoint that encapsulates all others, the god’s eye vantage turns out to be simply the human viewpoint brought to a recognition of its own partiality.

On Burnyeat’s reading, Heraclitus’ goal is just this recognition. The lineaments of the human view and its language-structured oppositions should remain what they were – we continue to live and think within it just as before. The one difference, now, is that we know – and have really taken in – the fact that ours is only one possible viewpoint among others.^[62]

Burnyeat’s analysis of Heraclitus both departs from and returns to another argument that juxtaposes the equal, opposed appearances that belong to humans and nonhuman animals – specifically, Sextus Empiricus’ version of Aenesidemus’ first ‘mode’ or argument for generating suspension of judgement.^[63] Sextus’ argument, which itself incorporates elements of Heraclitus’ human-animal juxtapositions, focuses on the kinds of appearances with which an object presents us on the level of sense perception.^[64] Sextus uses the differences in physiological structures and responses among animals as evidence that the perceptual appearances that a given object presents to nonhuman animals are likely to be in conflict with the perceptual appearances that the same object presents to humans. He infers from the diversity in modes of reproduction and in structures of perceptual organs to diversity and conflict in the perceptual appearances to which they give rise, and infers from diversity and conflict in preferences and aversions to diversity and conflict in the perceptual appearances that underlie them. Faced with these likely conflicts, we try to judge between the appearances in order to determine what an object is really like. However, as humans, we cannot pretend to be impartial – the perceptual appearances that seem most compelling to us will be the ones that we receive through our own sense organs. In recognizing our inevitable bias, we also see that we are not competent to judge between these different appearances, and must therefore suspend belief with respect to them.^[65]

Burnyeat is critical, however, of Sextus’ formulation of the argument, noting that there are situations when we do favor the sensory appearances that object presumably presents to a nonhuman animal over those that it presents to us – for instance, in trusting a dog’s sense of smell over our own while tracking. Our conception of “better and worse perceptual equipment” provides an empirical basis that allows us to judge between the perceptual appearances that belong to different species without (necessarily) opening ourselves to the “charge of anthropocentric partiality” with which Sextus’ argument culminates (Burnyeat 2012a, 309). Burnyeat frames Heraclitean god’s eye vantage as a more compelling formulation of that charge, which he thinks we will not so easily escape. According to Burnyeat, Heraclitus does not stop, as Sextus does, at problematizing our human partiality via the conflict between our sensible impressions and those of nonhuman animals. Instead, he challenges the very structures through which that conflict is recognized and assessed. Specifically, Burnyeat’s Heraclitus hones in upon the uniquely human character of language as a structure that pervades and conditions our experience – in particular, through the conception of opposition and contradiction upon which that structure rests. Again, this creaturely idiosyncrasy is visible through our inability to cognize unqualified opposites (e.g., ‘good’; ‘bad’) either together or independent of one another – an inability that reveals both our categories and the relationships between them to be idiosyncratically human.^[66]

My approach to the human-animal juxtapositions has certain affinities with Burnyeat’s analysis of the god’s eye vantage. Like Burnyeat, I understand Heraclitus to create a set of conditions under which the attempt to grasp the opposites together as a unity reveals the species-relativity of our categories. Like Burnyeat, I understand this challenge to our categories as bound up with a more general challenge to human faculties and human knowledge. On Burnyeat’s analysis, however, that challenge flows from the species-relativity of our language – in particular, the rules around opposition and non-contradiction that govern it. In addition, for Burnyeat’s Heraclitus, our species-relative quality ascriptions offer a partial, but genuine, insight into reality.^[67] On my reading of the human-animal juxtapositions, by contrast, the challenge to human categories flows not from human language as such, but from the species-relative values that it reflects – that is, the natural volitions (needs and preferences) that simply express a creature’s physiological conditions of survival or flourishing. Categories such as axiological opposites and (apparently) non-axiological opposites, as well as the objects in which these inhere, seem to be reducible to these species-relative volitions in such a way that they tell us nothing about the creature-independent object. As a result, on my reading of the human-animal juxtapositions, these species-relative categories lose even the partial insight into reality that Burnyeat finds in them.

The affinities between Heraclitus’ human-animal juxtapositions and Sextus’ argument from differences among animals are also suggestive. Both draw upon the physiological parameters that shape various creatures’ responses; both use the alignment and conflict between the responses of human and nonhuman animals as a challenge to the audience’s judgements. However, Sextus’ argument moves from physiology to the sense impressions to which it gives rise; he takes the conflicts among different sense impressions as the source of the challenge, rather than the fact that they arise from a distinct kind of physiology or nature. For him, this challenge is mediated by the demand to adjudicate between the conflicting sense appearances, and the inability to do so. Heraclitus, I have argued, moves from conflicting quality ascriptions to the nature or physiology from which they arise. For him, the crux of the challenge lies not in the opposed, aligned qualities themselves, but in the reduction of those qualities to their ground in a species-relative nature or physiology. For Heraclitus, this challenge is mediated by the demand to grasp the conflicting quality ascriptions together, and the inability to do so.

On Burnyeat’s reading, then, Heraclitus’ alignment of conflicting opposites attacks the partiality of human language – though ultimately allows that our language-circumscribed cognition still possesses some (limited) insight into reality. Sextus’ argument, by contrast, hones in on the sense impressions that belong to various creatures as a result of their physiological structures. He aligns conflicting human and animal perceptions in order to stimulate, neither acceptance nor rejection, but a suspension of judgement regarding those perceptions. Finally, Heraclitus’ human-animal juxtapositions, on my reading, target human-relative values, show how various distinctions arise from those values, and, finally, seem to demand that the audience reject the whole package. Yet, like the sense impressions in Sextus’ arguments, the creature-relative responses in the human-animal juxtapositions seem to arise automatically from species nature or physiology.

These points of overlap and divergence, finally, bring out the scope of the epistemic challenge in the human-animal juxtapositions – or, more precisely, the challenge that they retain even if their challenge to distinction-making is framed provisionally with the aim of goading the audience towards greater self-awareness. The juxtapositions seem formulated in such a way as to provoke, not merely epistemic dissatisfaction with, but visceral distaste for, the various creature-relative judgements. On a rhetorical level, Heraclitus leans upon the unpleasant associations attached to animals traditionally represented as ignorant and base. On an inferential level, he moves from the failed effort to affirm both of the creature-relative opposites to something closer to a denial of both. In this, the human-animal juxtapositions fit well with the general tone of a thinker who not only shows marked pessimism about human epistemic potential^[68] but also a perplexing tendency to blame humans for this species-dictated ignorance^[69] and even, it sometimes seems, for the very fact that they are human.^[70] While the human-animal juxtapositions, taken on their own, do not indicate what is to be done to remedy this natural-yet-blameworthy fault, they do indicate that even a partial solution will have to embrace both basic volitions and cognition, in a process that feels not only like a re-evaluation of, but a full-on attack upon, instincts of choice and avoidance, pleasure and displeasure, as well as their expression in differentiations between opposites or objects. As such, the juxtapositions may open up a new vantage on Heraclitus’ conception of philosophical method – as witnessed, for instance, in the emphatically negative path sketched in B18: “If one does not hope for the unhoped-for, we will not find it, for it is unsearchable and trackless” (ἐὰν μὴ ἔλπηται ἀνέλπιστον οὐκ ἐξευρήσει, ἀνεξερεύντον ἐὸν καὶ ἄπορον).

The verb ἔλπηται (hope, expect), with its cognate noun ἐλπίς, is at once cognitive, volitional/affective, and constitutional/natural. It refers to beliefs, particularly expectations of things whose outcome is uncertain, and to the attitude (hope, fear) with which one looks upon those things. Yet it also refers, especially in Archaic and early Classical poetry, to a deficit in human (qua mortal) nature, involving an ingrained mode of living by delusions.^[71] In light of these connotations, the problems of the ἐλπίς that guides and impedes the searcher in B18 harmonize well with the problem of the needs and preferences that guide distinction-making in the human-animal juxtapositions. In demanding that the audience conduct their search by expecting the unexpected and hoping for the unhoped-for, B18 describes the kind of radically transformative process through which they might continue seeking knowledge in the face of such obstacles – a process involving changes in the object sought, the means by which it is sought, and even, perhaps, in the basic characteristics or ‘nature’ of the one who seeks it.