CHAPTER 8
Parts and wholes in the Parts of Animals
Thomas K. Johansen
Introduction
PA I.5 (645a35-6) sends a strong message that the parts of the animal body are to be studied
for the sake of the animal as a whole. This means that the material parts are primarily of
interest for their contribution to the form of the whole living being. It is clear that this
contribution lies in the bodily parts´ instrumentality, their function in the performance of the
psychological functions that characterize the various animals. Yet Aristotle´s expository
method in the PA may seem to run counter to his holism. If, as Aristotle suggests (644a24), it
is the lowest or ‘indivisible’ species which are the substances, then (in accordance with
holism) we should study the parts of animals at this level (644a32). Accounting for the parts
of animals at a higher level would run the risk of abstracting from the explanatory
dependencies between body and soul that obtain at the level of the actually existing species.
Yet many of the parts of animals are common to several species, so explaining them for each
species, Aristotle says (644a36), would be repetitive and tiresome. We find thus in the PA
two opposed explanatory tendencies: one ‘upwards’ towards the more common and greater
simplicity and another ‘downwards’ towards the ultimate species and greater complexity. His
proposed solution is to account for the various bodily parts at a general level and to descend
to the species only when the parts differ significantly. In this paper I address some of the
difficulties that have to be resolved before one can accept Aristotle´s solution.
Essences and parts
In the first chapter of the PA, Aristotle recommends that we explain the parts of an animal
with reference to the animal´s essence:
T1 Hence it would be best to say that, since this is what it is to be a human being, on account
of this it has these things, for it cannot be without these parts. If one cannot say this, one
should say the next best thing, i.e. either that in general it cannot be otherwise, or that at least
it is good thus. And these things follow. And since it is such, its generation necessarily
happens in this way and is such as it is. (This is why this part comes to be first, then that one.)
1
And in like manner one should speak in precisely this way about all of the things constituted
by nature.’ (PA I.1 640a33-b4)1
The essence is the object of the definition. Here, the essence, ‘what it is to be a human being’,
explains the parts of the animals. As the passage is often read,2 this may be immediately
because the part is contained in the definition, as ‘wing’ might be part of a definition of ‘bird’
as ‘an animal with wings’. However, it may also be because the part is necessary for the
essence to be realized, even if it is not itself part of the essence. Here the notion of
hypothetical necessity captures the relationship between essence and part. So, we might say
that feathers are hypothetically necessary for the bird if it is to have wings. In either case,
however, the essence would be the starting point for explaining the parts of the animal.
A form of essentialism is, then, a presupposition of Aristotle´s accounts of the parts.
Importantly, the essence is not itself the target of explanation, the part is. We should therefore
not expect Aristotle in the PA to spell out definitions of animals. Rather what is called for is
to mention those features of the definition that are necessary for explanation. But this may
only be some features of the animal´s essence. His essentialism is therefore also compatible
with the absence from the PA of any full definitions of animal kinds.3
Essence, soul and body
The way in which the essence is the starting point for explaining the parts of the animals is
explicated in two complementary ways in PA I.1. The first is in terms of the distinction
between form and matter. Animals, like all natural beings, are composites of form and matter,
with the soul playing the role of the form. The nature of an animal is both soul and body. The
natural philosopher thus studies both soul and body.
However, priority goes to the soul as Aristotle says in this programmatic passage:
T2 it will be up to the natural philosophers to speak and know about the soul; if not all of it,
about that very part in virtue of which the animal is such as it is. He will state both what the
soul or that very part of it is, and speak about the attributes it has in virtue of the sort of
substantial being it is, especially since the nature of something is spoken of and is in two
ways: as matter and as substantial being. And nature as substantial being is both nature as
mover and nature as end. And it is the soul – either all of it or some part of it – that is such in
the animal´s case. So in this way too it will be requisite for the person studying nature to
speak about soul more than the matter, inasmuch as it is more that the matter is nature because
of soul than the reverse. (PA I.1, 641a22-32)
1
Translations of the PA here and throughout by J. Lennox, Aristotle. On the Parts of Animals, Oxford 2001
(2001a).
2
In line with the reading of the passage advocated by Balme, Cooper and Gotthelf, cf. Lennox 2001a: 134-5.
The point in T1 about the order of generation may support this reading: the parts that will develop gradually
towards the creature´s essence will clearly include parts that are essential to it as a certain kind of animal,
including nutritive, perceptual and, sometimes, locomotive parts.
3
Cf. Gotthelf 2012: 155: ‘It has often been claimed, and not without reason, that there are no definitions of
animal kinds in PA, or anywhere in the biological treatises.’
2
The natural philosopher will treat of both form and matter, soul and body, since both are
nature. However, the starting point will be the soul, since the soul is the final cause of the
body,4 that for the sake of which the body is, and so prior in explanation to the body.
Secondly, the priority of the form as the final cause is explicated also in terms of necessity.
The body is necessary, hypothetically necessary, for the functions of the animal, just as
certain materials are necessary for the axe if it is going to be able to split (PA I.1, 642a9-10).
Again, Aristotle gives priority to the form over the matter in this relationship: the matter is
necessary for the form (I.1, 640b27-29, I.5 645a30-36, 645b15–20, II.1,, 646b2-3) so the
matter is to be explained by its contribution to this form.5
There is a debate, explored in this volume by D. Henry and M. Peramatzis, as to
whether the essence that provides the starting point of the explanation of the parts is purely
formal, or whether the form itself involves reference to matter. I shall not engage in this
debate directly here.6 But I shall suggest that it is important to see the way the essence
explains the body in terms of levels of explanation. Here are two examples, nutrition and
perception to clarify the point. Nutrition serves the function of maintaining the animal. In DA
II.4 we are told that the soul was the efficient and formal cause of nutrition, understood as the
process whereby the animal cooks up the food to provide the final nutriment for the body.
The account of nutrition was functional, making no reference to any particular living being,
nor therefore to any particular kind of body or food or any particular kind of final nutriment.
The account of nutrition is ‘purely’ formal: there is nothing here in the account of nutrition
that corresponds to the material account of anger as ´boiling of the blood around the heart´
(DA I.1).7 It is in the PA and the GA we learn that the final nutriment for some animals is
blood, for others something analogous. We learn how the cooking works in terms of heat and
we learn how blood works in relation to the flesh. A different story applies to plants, though
this falls outside the brief of the PA. We may say that the account of the DA was formal and
shied away from any physiological detail because Aristotle wanted here to articulate an
account of nutrition at its most general level, an account which would apply to all living
beings. In the PA, meanwhile, we are interested in the bodily parts of different kinds of
animals so we shall now account for the psychic functions as realized in particular kinds of
matter. So we explain blood by its role in nutrition. Talk of a blooded animal will inherit,
however, the functional understanding of nutrition from the DA: blood is final nutriment of
the sort that will support the animal´s continuation as a certain kind of ousia. Whatever we
say about the different material parts of different animals will be said against the background
of the general formal account of nutrition given in the DA.
4
De Anima II.4 provides the subtext to this passage.
Cf. ‘Nature is of two sorts, nature as matter and nature as form, and the form is the end, and since everything
else is for the end, the form must be what things are for.’ (Phys. II.8, 199a30); ‘There are four causes: first, the
final cause, that for the sake of which a thing exists; secondly, the formal cause, the definition of its essence
(and these two we may regard pretty much as one and the same).’ (GA 715a4-6)
6
I argued in The Powers of Aristotle´s Soul (OUP 2012) that when defining the soul the form should be
specified by the natural philosopher in such a way that it would support an account of the body by hypothetical
necessity.
7
This is not to say that it is not implied that concoction is a materially realized process but rather that there is no
explicit mention of what kind of matter realizes the process.
5
3
Now when we proceed to use blood as a differentia in accounting for other parts of the
animal, we should not think that the explanatory connection with the formal account of
nutrition has been severed: it is just that what is of immediate relevance to the part that we
are trying to explain may be that the animal has already been determined as blooded. So
having blood will also explain why animals have fat, the result of a rich diet not having been
absorbed by the flesh from the blood (PA II.5, 651a20-27). Fat in turn may explain infertility
in some animals (651a12-17). Bodily differences with an explanatory origin in basic
psychological functions may help explain other bodily differences.8
A parallel case is presented by flesh. All animals qua animals must have perception,
of which touch is the most basic modality. The flesh is composed as the primary organ of
touch. Being composed of the four basic tangibles, the hot and the cold, the wet and the dry,
the flesh bestows on the animal the ability to be affected by these qualities. Again, we may
say that flesh (or an analogue) is part of the nature of an animal as such, and the natural
philosopher will offer an account of this part both in terms of the psychological function,
perception, and the flesh. The flesh will in this way stand as a differentia of animals from
non-animals. The flesh in turns works as an explanation of the animal´s other parts. So being
soft the flesh requires supportive structures, typically bone, which in turn will help explain
the sinews. However, the fact that flesh explains bone does not mean (though it does not
exclude either) that bone or sinews, in turn, is part of the definition of the animal kind in
question. Again, it is important to keep in mind the explanandum: the various animal parts.
To explain them we require differentiae that lead us back to the essence of the animal, but we
are not thereby using these differentiae to define the animal. This is not the target of the
exercise.
It should come as no surprise therefore that matter is referred to in the differentiae
explaining the various parts. The starting point will ultimately be a differentia in function or
end, like nutrition or perception, but such functions require bodily parts which themselves are
explanatory of further differentiae. However, as we shall see, the differentiae that are referred
to in the accounts of the parts are of a wide variety, much wider than just a material
composition, like flesh or blood. Some refer to a function, others to a shape or size, yet others
to a life-style, yet others again to the material composition.
I Levels of explanation
Now the explanatory priority of an animal´s essence in relation to parts might suggest that the
parts of animals should be accounted for at the level of the indivisible or ultimate species, cat
8
I focus here on blood, but much more could be said about the other aspects of nutrition, for example
concoction. For a detailed analysis along the lines I am suggesting, see Gotthelf 2012: 165-171. He sums up the
explanatory principles involved as follows: ‘the fact about animals which distinguishes them from plants, and is
central to their mode of nutrition (their ‘threptic’ soul), is that they concoct their food inside themselves. The
argument thus starts with a basic fact about all living things, then adds a basic fact about (in effect) animal soul,
then derives the need for certain organ-systems. All this is a general framework within which existence of the
stomach and associated organs is explained, as the organs which perform these necessary functions’. (169)
4
or dog, rather than higher kinds, mammal or animal.9 The following argument, based on
statements from PA I.4-5, would support this suggestion:
1) the matter of an animal should be explained in relation to the form that most properly
defines the animal, e.g. what it is to be a human being (I.5 645b-36);
2) parts out of which an animal is composed are its matter (I.5 645a25-36);
3) the form that defines the animal is the form of the indivisible species, e.g. human being
(I.4 644a29-32);
4) the parts of the animals should be explained in relation to the form of the indivisible
species.
So, the parts of animals should be discussed at the level of generality where the animal is
most properly defined, which would mean the level of the lowest species, crow, sparrow or
crane, say, rather than at a more general level such as bird or oviparous.
However, Aristotle also raises a procedural difficulty:
T3 Since, however, it is the last forms that are substantial beings, and these, e.g. Socrates and
Coriscus, are undifferentiated in respect of form, it is necessary either to state what belongs
generally first, or to say the same thing many times. And things that belong generally are
common; for things that belong to many we call general. There is, however, a puzzle about
which of these two should be our subject. On the one hand, in so far as what is indivisible in
form is a substantial being, it would be best, if one could, to study separately the things that
are particular and undivided in form – just as one studies mankind, so too bird; for this kind
has forms. But the study would be of any one of the indivisible birds, e.g. sparrow or crane or
something of this sort. On the other hand, in so far as this will result in speaking many times
about the same affection because it belongs in common to many things, in this respect
speaking separately about each one is somewhat silly (or ‘absurd’, hupatopon) and tedious.
(PA I.4, 644a23-b1)
Aristotle sees a tension here between two explanatory approaches to the parts of animals.
9
For the notion of ultimate species see D. Henry in this volume. He defends two claims:
(1) The epistemological claim. The aim of Aristotelian zoological division is to identify and organize the
essential features specified in the definition of each indivisible species of animal.
(2) The ontological claim. Those indivisible species constitute the ousiai of Aristotle’s zoology.’ While I agree
with his defense of (2), I am less convinced of (1). Henry argues that ‘Since the object of definition is always a
substance, it stands to reason that in the zoological works animal parts constitute the substances “in the primary
and strongest sense” (106). Indeed (as the title of his paper suggests) Pellegrin argues that Aristotle’s zoology
has no need for animal species. I will return to Pellegrin’s ontological thesis in the next section (animal parts
constitute the real substances of the living world). In this section I want to focus on the epistemological claim
that, according to PA I.2-3 at least, the true objects of definition are “the parts, their properties, and their
variations” and not the animal species themselves (op. cit.). By contrast, I shall argue that, on the most natural
reading of PA I.2-3, the ultimate objects of definition are the indivisible species of animal to which those parts
belong.’ However, the fact that PA I.2-3 presents the method of division of animal species by parts as
differentiae does not mean that the parts are accounted for to define the species. Rather Aristotle may mean to
justify why the differentiae are the basis from which to account for the parts: they define the species. The
essence is the explanatory starting points for the parts in that its differentiae will be referred to in the account of
the parts. This alternative explanation is more readily compatible with two features of the PA: both that other
differentiae are referred to frequently, and that not all the differences are defining differentiae.
5
On the one hand, there is a preference, as we have seen, for an account of the parts at the
level of the indivisible species, the so-called lowest or infimae species. Aristotle´s general
view is that these species are ontologically prior to more common kinds. So, in DA II.3
(414b20-28) Aristotle says that 'it is absurd to seek a common account (logos) in these and
other cases (sc. like geometrical figures and souls) which is not a proper account of any of the
things that are and is not in accordance with a proper and indivisible kind (eidos), while
leaving out such an account.’ (414b25-28). While Aristotle is not saying in DA II.3 that there
should not be a common account, the suggestion is that a common account is secondary in
relation to an account of the indivisible species since this is the level of things that are.
To illustrate the point, imagine zooming in with your binoculars on a bird in the
distance: at first you see features common to all birds, a wing, a beak, little by little features
common to corvids, until a clear picture emerges of crow. The other features were there in
virtue of this crow. It is birds at the level of specificity of a crow that exist, the more
undifferentiated features you saw at first are only there because there are crows, or jays or
sparrows. Dividing is like zooming in: only by homing in on the lowest species do we reach a
point where the universal corresponds to real-world entities. If we insist that our accounts of
parts should be real, then the accounts at the level the infimae species would seem to have the
strongest claim.
The countervailing consideration suggested by T3 is this: if we begin from the level
of individuals, like Socrates and Coriscus, it is clear that they will have most things in
common, indeed everything which qualifies them as human beings. There is no need then
describe these features relatively to each individual. But by the same reasoning one might
think that, since human beings share many features with other land-animals or blooded
animals, those features should also be described at a higher level of generality than that of
human being. Considerations of simplicity would seem then push us higher up towards the
more general accounts.
Add to this that Aristotle says not only that it would be boring to go through the
lowest species one by one, but also that it would be ‘somewhat absurd’, that is, I take it,
lacking in rationality. A likely reason for this may be that there are features of the lowest
species that belong to them not insofar as they are that species, but insofar as they belong to a
larger kind. So a human being has a heart not qua human being, but qua blooded animal, as
all blooded animals have a heart. ‘Blooded’ is the term that explains why humans have a
heart.10 Compare on this point Aristotle APo I.5, 74a26–33: ‘if you tried to demonstrate
separately for each kind of triangle that each has internal angles equal to two right angles
‘you would not know [this], except sophistically . . . for you do not know [that it has two
right angles] as a triangle, nor do you know it of every triangle, except in a numerical sense’.
The explanatory middle term is known as the ‘commensurate universal’ in APo I.4. Similarly,
to explain why the lowest species of animals have a certain part you may need to ascend to a
feature of a more common kind. It would be less rational, and ‘somewhat absurd’, to give the
account here at the level of the ultimate species.11
10
Referred to as the ‘commensurate universal’ in APo I.4.
The point is made by Lennox (2001a) 170 and developed in Johansen 2012: 64-6. See also Lennox 2001b:
29–30. Henry (2021) is skeptical of this connection with the APo: ‘(Aristotle) need not have anything more
complicated in mind here than the simple point that it is more efficient to give one explanation for why all birds
11
6
We have then considerations for and against accounting for the parts of animals at the level
of the indivisible species. On the one hand, it seems that the parts of animals should be
explained at the level of the ultimate species since this is the most proper level of definition,
and so the level at which we can best see how the parts contribute to the substantial being of
an animal. On the other hand, many parts are common to many species of animal and are
realized in them in virtue of an animal´s more generic attributes and not because it is this
particular species of animal.
In PA I.4 Aristotle moves to resolve the issue:
T4 Perhaps, then, the right course is this. In some cases – whenever kinds are spoken of by
people in a clearly defined manner and have both a single common nature and forms in them
not too distant – we should speak in common according to kinds, like bird and fish and any
other there may be that, though it is unnamed, embraces, like a kind, the forms within it. But
whenever they are not such as this, we should speak one by one, e.g. about mankind and any
other such kind. (644b1-7)
We should speak of the parts together where they share a single common nature as do species
of birds, but where the distance is greater, as between birds and fish, it seems that separate
accounts are appropriate. The key point is whether there is a single explanatory feature, a
common nature, that applies to the species. Where the species are too distant in not sharing a
common nature, it is reasonable to think that the cause of their parts would not be of one kind
or nature. Here the only way of bringing these different parts together in one explanation is
by analogy. But, as we know from the APo I.10 (76a38-42), analogy has a lower explanatory
value because the explananda do not fall under a shared genus.12 So wings and fins may
have similar functions in birds and fish, but you cannot demonstrate facts about birds from
facts about fish.
Even if wings and fins are common by analogy why should there not be common
explanations of them? They both perform the job of propelling the animal forward in their
respective media. Admittedly, unlike perception, forward locomotion is not a function that
birds and fish share qua animals, as shown by the fact that not all animals move. Yet they still
share a highest genus, animal, so there is no reason to think that forward motion in the two
cases would be ambiguous.13 Why then insist on mere explanation by analogy? One answer
might be that the shared function is specified at such a high level of generality that it has no
specific consequences for the sort of material part that would serve it. We are after all, as
have split wings than to repeat that explanation over and over.’ (95, n.27). I take it, in contrast, that it is
‘somewhat absurd ‘in the sense of less than rational to speak of individual cases when the part ‘belongs in
common to many things’. But this point seems exactly to trade on the question of the level of generality at
which an attribute properly belongs.
12
See Lloyd (1996) 142-8 on the relationship between analogy and Aristotle´s demonstrative method in the
APo.
13
Lloyd (1996) 142 says that ‘The feathers of birds and the scales of fish are not the same in genos in one sense,
but they both fall perfectly satisfactorily within the purview of an inquiry into animals, where animal provides
the determinate genos to meet the APo condition.’ The question is whether it is sufficient to have a shared genos
‘animal’ to provide an unambiguous explanatory term for the protective covering of fish and birds. Aristotle´s
emphasis on the distance of the two kinds – to be sure within the genos ‘animal’ – suggests that it is insufficient.
7
emphasized in PA I.1 (640a33-b4) interested in explaining the different bodily parts of
animals from the point of view of their forms. The point, as we saw, is developed by
reference to the notion of hypothetical necessity. Aristotle envisages the formal-final cause to
provide hypothetically necessary explanations of the material parts (642a32-642b4). But, if
so, I am suggesting, we may need to descend to the level of fish or bird form before we can
see why this form requires a certain bodily realization. The description of functions at too
high a level of generality would not support such entailments.
To give one example, for an account of function to support an explanation of the
bodily part as hypothetically necessary, we may need to factor in facts about how the animal
lives, for example whether it lives in water or air.14 Simply saying that an animal is able to
move forward will have only vague implications for the nature of its organs of propulsion. In
contrast, forward propulsion in water will require a specific kind of body part, fins, as will
progression in air, wings. But to bring in the environment is already to descend to a notion of
function specific to these kinds of animals living in water or air.15 Put differently, the level of
specificity at which an account of the function ties up with an explanation of the animals´
bodily parts is lower than the level at which the account applies indifferently to water and
airborne animals, that is, we need to descend from the level at which they are one by
analogy.16
There seems, on this interpretation, to be a justification for not giving common
accounts of parts belonging to distant animals.17 We can treat kinds that are relatively close
as one, such as various birds which have a number of joint features as birds, while treating
those separately which are most distant from each other, like birds and fish, or man, who in
some ways occupies a special position in relation to all the other animals.18 The distance in
kind between the animals is reflected in the explanatory stance we take to their parts.
Whether they differ at a high level of generality and are similar by analogy or whether they
are the same kind and differ only in terms of degree,19 ‘kind’ is here indexed to kind of living
being.
T5 It is necessary first to divide the attributes associated with each kind that belong in
themselves to all the animals, and next to try to divide their causes. Now it has been said
14
Cf. PA III.4, 665b2-5: ‘Just as, with the external parts, there is not a use provided to all animals, but rather a
distinctive provision has been made for each of them related to their ways of life and movements, so is it natural
that the internal parts are also different in different animals.’ See further Lennox (2010), who argues plausibly
that ‘an animal’s many parts and corresponding activities are provided with a unified explanation by identifying
their contributions to the one, or ‘way of life’, of that animal’ (333). From the point of view of my argument
here, the relevance would be that information about the way of life is crucial to generating the hypothetical
necessities that explain the particular parts of an animal.
15
Aristotle refers to the seal having rear feet that are ‘entirely fish-like’, like fins (IV.13, 697b4-5), although the
seal is a ‘dualizer’, in some respects a water-animal, in others a land-dweller.
16
Aristotle in general takes it that natural philosophy (as opposed e.g. to mathematics) should account for the
form in a manner supports hypothetical necessitation of the matter, see Phys. II.2 and DA I.1 (cf. Johansen 2012:
147-57).
17
It is worth noting that functional analogies for parts of the body sometimes extend outside living beings, for
example to irrigation systems (the blood vessels), anchors or nails (the viscera, III.7, 670a10,14) and a hearth
(the heart, III.7 670a25).
18
See, e.g. PA I.1, 645b5-10, Lennox 1999.
19
For some cases where Aristotle does not seem to apply the distinction consistently, see Lloyd 1996: 150-152.
8
before that many common features belong to many of the animals, some without qualification
(such as feet, wings, and scales, and affections too in the same way), and others analogously.
By analogously I mean that while some have a lung, others have, not a lung, but instead
something different which is to them what a lung is to those that have one; and some have
blood, while others have its analogue, with the same potential that blood has for the blooded.
To speak separately about each of these animals as particulars, as we also said before, will
result in saying the same things many times, whenever we speak about all the attributes; the
same attribute belongs to many animals. (I.5, 645b1-14)
T5 reiterates the message of T4 that we can speak generally of parts that belong to many
animals to avoid repetition.
However, one might still wonder on this account why the distance between kinds
should by itself translate into a distance between the parts of animals of those different kinds.
So one might in principle think that a bird and a fish could have some fundamental
difference, say that one flies and the other swims, which would determine one kind having
wings and the other fins, while taking it that they also share a large number of very similar
parts. So while wings and fins might be the same only by the weaker relation of analogy, the
other parts could differ by the stronger relation of the more and the less. In other words, there
is no reason why a particular difference between two kinds, basic as it might be to
differentiating them as kinds of animals, should translate into an equally basic universal
difference between all the parts of those two kinds. One could point to the variety of
differentiae that Aristotle uses to characterize the parts of animals. So, a fish is a water
animal, as well as blooded and oviparous, while a bird is a flying, blooded, two-legged,
oviparous animal. This in turn would suggest that some features of fish and birds will be
common qua blooded, and so may differ by the more or less, while the means of propulsion,
reflecting a more fundamental difference, will only be the same by analogy.
The concern is then that the relative difference in kinds of animals does not allow us
consistently to distinguish when their parts are the same merely by analogy or by the more or
the less. Now there are at least two things to say in reply to this worry. One is that Aristotle
sees the differentiae as closely interrelated. It is not the case that a fish operates as a blooded
animal in isolation from its nature as a water animal. As Aristotle says,
T6 Yet if it is impossible for some indivisible and unitary form of substantial being to belong
to animals that differ in form – rather, the form will always have a difference, as bird differs
from mankind (for their two-footedness is other and different) – then even if they are blooded,
either their blood is different, or blood should be reckoned as no part of their substantial
being. (643a1-5)
So the options are either to treat ‘blooded’ as different in the two animals that differ in form
insofar as ‘blooded’ is a differentia or to say that ‘blooded’ is the same but does not enter into
the definition of what the thing is. The latter alternative seems less attractive because the
contrast between ‘blooded’ and ‘unblooded’ is a fundamental difference between kinds
across the PA. Yet the former option might seem to open the door to homonymy in the terms
that differentiate the kinds. If ‘blooded’ differs when it is a differentia of dog and sparrow,
representing different kinds of being for dog and sparrow, what it is to be ‘blooded’ in these
9
two cases must surely be something different. But if the being of ‘blooded’ differs we seem
to have a case of homonymy, that is, one term with two different definitions.
However, there is no need to draw this conclusion, as we can see from Aristotle´s
account of two-footedness in humans and birds. Certainly, the two kinds are two-footed in
different ways: ‘[birds´] legs are two, as with mankind, thought bent inward, as with fourfooted animals, and not outward, as with mankind’ (IV.12, 693b2-3). However, there is no
ambiguity in terms of their being animals with two feet. In the same way, one may say that
the two kinds are blooded in different ways, but there is no ambiguity in terms of their being
blooded animals. The blood of these animals is in both cases a concoction providing
nutriment produced by heat providing nutriment to the body, just as the two feet of both birds
and humans support them and propel them forward in fundamentally the same manner. These
are not cases of homonymy such as the living man and the man in a portrait or the living body
and the dead body, where the difference does not arise from a specification of a shared
functional property. The more general point here seems to be that differentiae allow for
modification at the level of the lower kinds. As long as we allow for the different ways of
being blooded as a specification of a more general way of being blooded, there is no need to
worry about homonymy.
Aristotle´s point that the parts of animals differ by degree presupposes that they are
said without homonymy: it makes no sense to speaker of one part being more less than
another – say one animal having harder or softer skin – unless one is comparing them in the
same respect. For the same reason that we may differentiate the parts by their degrees we
may also distinguish without homonymy the kinds by the degrees to which their parts are
qualified by a property: some animals are more warm-blooded than others, for example.20
II Parts and holism
We began by asking whether Aristotle´s procedure of discussing parts at a general level
above the lowest species was consistent with the claim that it is the lowest species that are
real and the fundamental objects of definition. So far this procedure has appeared justified.
But it is worth considering the issue in more detail from the point of view of Aristotle´s
holism. It may seem here that the various explanatory factors are so interwoven that the
appropriate level of accounting for living beings must generally be that of the ultimate
species. This would be because the parts of the organism are so closely related to each other
that it seems that they can only be understood if we consider the full range of explanatory
factors, and this would inevitably bring the explanation down to the level of the indivisible
species.
We start by observing the functional interconnectedness of the parts. The functions of
the parts of animals form a hierarchical structure with a common point at the top, the essence
of the animal. We see this clearly in PA II.8-9, where Aristotle treats the flesh and the bones.
Flesh is a particularly central example since all animals must have it, or an analogue, to
possess the basic capacity that defines them as animals, perception. Flesh is the organ of
20
On the importance of ‘the more and the less’ in differentiating the animal kinds, see Lennox 2001c: 166-7.
10
touch, without which there can be no perception. Moreover, the flesh is basic for the
composition of the whole body in the sense that all the other uniform parts – bone, skin,
sinews, blood vessels, hair, etc. – are for the sake of the flesh (653b30-33). These parts serve
to support or protect the flesh. Aristotle has then a general functional conception of flesh, as
an instrument of touch, which applies to all animals, and he has a corresponding general
functional conception of the other uniform parts as preserving the flesh. This function allows
us, unambiguously, to fix a single reference not only for ‘flesh’ and ‘its analogue’ but also for
the subservient parts. So bone is a hard structure for ‘the preservation of soft tissue’, with
fish-spine or cartilage as the analogue parts in other animals (653b32-35). From this common
starting point, the flesh, Aristotle now differentiates between the parts on the basis of how
they preserve the flesh or its analogue in different animals. So, some animals have bones
inside, others (like crabs) on the outside. The explanation for this difference is that the latter
animals have relatively little internal heat and therefore need the covering on the outside to
preserve it, like a lid on a saucepan. Other animals, like insects, have neither bone nor soft
flesh, but a substance of a nature in between flesh and sinew, which provides sufficient
structural integrity for them not to require the additional support of bone (654a8-19).
Different animals have the function of touch, then, but have different parts to support
this function. We might say that while their parts are compositionally different, their
composition presents different solutions to the same problem of how to protect the animal´s
basic perceptive part. While there is no reason in principle why one function should not be
realized by a part with the same constitution in two kinds of animals, we find that the organs
vary according to a range of other constraints to do with their aims, ways of living,
environment, size, shape, and weight, available material, etc. Explaining how a part is
actually composed requires us not just to consider how perception in general might be best
realized but how it might be best realized for a creature which is subject to all of these
constraints.21 Given the complicated matrix of factors involved, such an explanation, we
might expect, will naturally tend ‘downwards’, away from general facts about animals or
their most common kinds, towards the lowest species. For each factor will introduce further
complexities that bring us closer to the sort of animal that actually exists, the ultimate
species.22
A good illustration is when the part serves two or more functions. Here we often need
to look at the ultimate species to understand how the functions are combined. So, Aristotle
says in PA III.1:
T7 All those animals that breathe and cool themselves from outside have the nature of the
mouth both for the sake of these functions and for respiration besides. For nature, in virtue of
itself, as we just said, puts the parts common to all animals to many distinctive uses; for
example, in the case of the mouth nourishment is common to all, while strength is distinctive
to some and speech to others, and again breathing is not common to all. But nature has
21
As Lennox 2001a: 335 observes, ‘the organs are constructed out of the various uniform parts, which are
constructed out of the blood, what is optional is the precise choice of organs to be constructed, and thus the
shapes, sizes, and combinations of uniform parts to be made into particular locations. It is the functions required
by the organism´s way of life that put constrains on this process, constraints which define the actions of the
formal nature.’ I take it that Lennox by ‘optional’ here means something like ‘variable’.
22
See further Henry 2021: 86, who rightly emphasizes the point.
11
collected all these uses together in one, producing a differentiation of this part for the
differences of its operation. That is why some mouths are narrower, some wider: those which
are for the sake of nourishment, respiration, and speech are narrower, while of mouths which
are for the sake of protection, all that are sawtoothed open wide. For since their strength lies
in their biting, it is useful for the opening of the mouth to be large; for the mouth bites with
more teeth and over a larger area to the extent that it opens more widely. (662a18-23)
Many body parts serve several functions. From the point of view of ingesting food there may
be no difference between the sort of mouth two species need, but combine ingestion with
speech, and together the two functions will demand that the human mouth be narrow, while a
dog, in contrast, needs a mouth that opens wide when it bites. Where a part has more than one
use, Aristotle in some cases refers to one use as ‘common’ (koinê khrêsis) and others as
distinctive according to kinds. So, ingestion is the common function of the mouth, while
different animals will employ it for various other purposes. Similarly, tasting is the common
function of the tongue in all animals, but some insects also use it as a sting for defense (IV.7).
The additional secondary function determines that the organ will differ in different animals
despite its common function. Each body part is here to be considered from the point of view
of a matrix of functions characteristic of the kind.
The combination of functions illustrates how parts are composed to serve the needs of
specific kinds of animals. If an animal needs both to taste and produce speech it requires a
tongue of a certain sort which will offset it from that of an insect, say. While the combination
of sound-production and tasting are not unique to humans – birds too are vocal and so require
broad and flexible tongues (II.17, 660a34-6, cf. HA IV.9, 535b1-2) – we might say that the
increased need for articulation of sound in a rational animal which learns by logos (De Sensu
I.1, 437a10-15) positively requires this kind of tongue for a human being.24 The elephant
provides another such case where we need to descend to the lowest species to understand
why the trunk is composed just as it is. 25
We have seen how the combination of functions constrains the composition of a part.
But the list of determinants is much longer. Sometimes the animal´s size and weight dictate a
certain arrangement of the parts. So the size of the horse and of certain oxen requires them to
have a bone in their hearts, while the weight of the elephant requires it to have solid hoofs.
The environment is a key factor in other cases. So fish are generally saw-toothed as their
watery environment requires them to cut up the food fast (III.1, 662a2-15). For the same
reason they have no or only a short tongue: the water that enters the mouth gives them little
opportunity to taste the food (PA IV.11, 690b24-31).
Further examples abound when we turn to parts that support other parts or
compensate for the lack of other parts. So animals with a heart have a liver, to aid its
function, while animals with a blooded lung have a bladder to deal with the excess moisture
that comes from drinking a lot. Other parts are present or composed in specific ways because
24
As Lennox (2001a) 149 notes the comparison between animal parts and artefacts ‘may suggest a view
congenial to modern “functionalist” theories’. Eye material will vary from one class of organism to another (cf.
PA II.10, 657a25-658a10), but such variations are related to differences in the organisms’ lives which require
differences in function, structure, and material.’ On the point see also Johansen 1997: 283-7.
25
The elephant´s trunk is perhaps the most celebrated case of one part with many functions, cf. PA II.16,
658b32-659b35 and Gotthelf 1997.
12
the animal lacks other parts that an animal of such a kind might be expected to have. So
Aristotle says that flies and bee-like insects have long forelegs to remove obstacles in front of
them, thereby compensating for their lack of clear vision (IV.6, 683a26-31). Another
example is the bladder of the tortoise: a hard-scaled animal would not normally possess a
bladder, but because the tortoise has a large lung and so takes in much liquid, which it cannot
secrete through the skin, it requires a bladder (III.8, 671a15-26). Conversely, some animals
lack a certain part because they have another part that already does the job for them. So
many-toed animals have no horns because nature has provided them with other means of
protection, such as claws or teeth (III.2, 66b30-34).26
These are different examples of how animals need specific parts to serve the totality
of their functions. Their explanations are final-formal not in the sense that they do not
mention material parts, but in that that those parts are viewed from the point of their enabling
or constraining the functions of the animal´s other parts. However, there are also
explanations which explicitly refer to a material cause as a complement the final-formal, as in
the case of the human hair:
T8 With respect to the head, mankind is the most hairy of animals, from necessity, on account
of the moistness of the brain and on account of the sutures (for where there is much moisture
and heat there must be much growth), and for the sake of protection, so that it may provide
covering, warding off the extremes of both cold and heat. And since the human brain is the
most moist, it is also most in need of this protection; for what is moist boils and freezes most
easily, while what is in the opposite state is less easily affected.27
Here the final-formal causal and material cause line up: the hair is both hypothetically
necessary for protection and necessary because of the moistness of the human brain.
Sometimes, however, the matter works in the absence of a final-formal explanation. Why do
four-footed, egg-laying animals not have upper and lower eyelids? Because scaly skin is too
hard to shape into eyelids (PA II.13, 657b11-14). Of course, there is a background final
causal explanation for the scaly skin: it serves the animal´s protection. But relative to the
explanandum, the absence of eyelids, the explanation counts as material. Similarly, hardskinned animals with dry flesh have no epiglottis, because this matter cannot be shaped into
this part. Instead, their larynx opens and shuts to prevent the food slipping down the wrong
way causing the animal to choke (III.3, 664a35-665a5).
Sometimes the material explanation is not determined, as in the manner of scaly skin
or dry flesh, by the fact that the matter has already been co-opted for another function, but
simply by there not being enough of it to go around. So the bear does not have much hair on
its tail because all the hairy material has already been used on covering the rest of its body.
‘For nature everywhere gives to another part what it takes from elsewhere’ (II.14, 658a36-7).
Also the material explanation here requires us to consider facts about the species as a whole.
26
On the principle ‘nature does nothing in vain’ see Lennox 1997.
Similarly for the viscera: ‘So, then: what each of the viscera is for the sake of has been stated, and they have
come to be of necessity at the internal limits of the blood vessels. That is, it is necessary that a bodily fluid
diffuse, and that this one, from which the body of the viscera come to be when it becomes constituted and
solidified, be bloody.’ (673a22-b1)
27
13
We would expect the final causal explanations to be holistic since, as we saw, the various
ends and uses of the body parts connect both with each other and ultimately with defining
features of the animal. We saw this point amply illustrated in the case of the flesh and the
various structures that followed upon flesh: an animal was defined by touch, which required
flesh and from there a range of supportive structures that worked out differently according to
the animal´s other needs given its environment, way of life, etc. However, it is noteworthy
that also the material causal explanations we have looked at are holistic, drawing in each case
on other facts, down to the availability of matter given its distribution to other parts of the
body. The final and the material causal explanations we have looked at are all holistic
focusing on what is necessary, possible and best for the animal in question.
The downward pressure of holism towards explaining the parts at the greater level of
complexity and so also of specificity does not stop at the ultimate species. For Aristotle
draws in also differences in sex and age within the lowest species. So, the female deer does
not possess horns, while the male does (III.1, 661b32-662a2). There is a final cause here in
terms of usefulness: the male is stronger and has greater use for the horns. Children, like the
lower animals, are described as ‘dwarf-like’ because of the disproportionate size of their
upper body (IV.10, 686b6-25). While these observations about differences within the same
species in a way go beyond the brief of the PA, accounting for the parts common to and
different to different kinds of animals, there is a sense in which they lie in extension of the
brief. For as in the case of children the feature is one that matches a difference between kinds
of living beings and have a shared cause. Female deer are brought in to illustrate a general
point about animals that nature only provides them with tools that they can use, while
children are mentioned in the context of discussing animal kinds that are dwarf-like. One
might say that intraspecies differences are adduced to confirm general causal points about the
species rather than for their own sake.28
III Holism and indivisible species
Aristotle recommended in T4 that
Whenever kinds are spoken of by people in a clearly defined manner and have both a single
common nature and forms in them not too distant – we should speak in common according to
kinds, like bird and fish and any other there may be that,… but whenever they are not such as
this, we should speak one by one, e.g. about mankind and any other such kind. (644b1-7)
How does this recommendation fare in view of the holistic accounts we have considered?
One might say that the individual references to indivisible species that we have come across
do not by themselves contradict Aristotle´s claim in T3; they just qualify as those cases
where we cannot speak of the parts of a kind in virtue of the features that all animals of a
28
As S. Connell suggests to me, this could help explain, to some extent at least, the collection of such subspecific differences in the HA, which D. Balme took to mean Aristotle was not an essentialist, cf. Balme 1987:
298.
14
more common kind possess. But this would not do justice to the holistic pattern of the
evidence. For the pattern does not suggest so much that the individual species provide
exceptions to be separately accounted for, but rather that the interconnectedness of the causes
of animal parts is such that an adequate explanation of a part generally tends towards the
level of the indivisible species.
It could be objected that the interplay of many causal elements is not a problem for
generality if the contribution of each of these causes can be understood generally. So, to use
an analogy, one might say that one complicated structure built out of Lego bricks can be
accounted for in terms of the same bricks as another complicated structure: the bricks
themselves are the same. Similarly, if each of the differentiae can be understood
independently across the species, then any combination of these differentiae, even if unique
to that species, can be understood as a combination of those modules.
However, as we have seen, causes do not work in the manner of Lego bricks or
Democritean atoms. Parts of animal bodies are not modules that are the same across different
structures. Animal bodies are not compounds of identical modules. The differentiae are not
the same across the animals they differentiate. Aristotle indicated that ‘blooded’ would differ
according to the kind of animal involved. Earlier I suggested that indexing ‘blooded’ to a
particular species need not be an obstacle to talking about it also in more general terms, as
long as we saw blooded in each species as a specification of a more general attribute.
However, while this remains true, the upshot of the last section is that it is the interplay of the
explanatory factors that determines the parts of an animal. We can only work out the parts of
an animal by considering how all the differentiae work together.
This point goes back to Aristotle´s anti-Platonic approach to division: we cannot
arrive at an explanation of an animal´s constitution by beginning with a simple differentia
(PA I.3 644a6-11). We need to keep all of a number of differentiae in play to understand how
particular parts are worked out. To give one final example, we see how the conjunction of
three differentiae determines the constitution of the lung:
T9 The lung differs in many ways in animals. Some have a blooded and large one, others a
smaller and spongy one. The live-bearing animals, on account of the warmth of their nature,
have a large one with much blood. The egg-layers have a dry and small one, though it is able
to open wide when being expanded, as for instance with the four-footed, egg-laying landdwellers, e.g. the lizards, tortoises, and all such kinds, and in addition to these the animals that
have a winged nature and are called birds. For the lung of all of these animals is spongy and
foam-like; and in fact when foam is stirred a large amount becomes small, and the lung of
these animals is small and membranous. (III.6, 668b24-33)
Four-footed, egg-laying, land-dwelling are here the differentiae that pick out animals, like the
lizard, which have small lungs capable of expanding.29 However, this combination of
differentiae is not necessary for having this kind of lung since also birds have it. Being
oviparous is the common differentia of the lizard and the bird here. Being oviparous in turn
correlates with being less warm than the live-bearing animals, which require a larger blood29
Yet a further consequence of having this sort of lung for four-footed egg-layers is that their spleen is ‘small,
firm, and kidney-like’ (III.7, 670b13-14).
15
filled lung. Yet being oviparous is clearly not on its own sufficient to pick out animals with
this structure, as there are other oviparous animals (insects, for example) that lack a lung, let
alone this kind of lung. It seems necessary therefore to involve a range of differentiae to
circumscribe the relevant kinds even though there is one differentia which in this context is
particularly explanatory of the part.30
To understand then why a specific part is worked out we need in principle to survey
all the differentiae of the animal kind. The challenge before us is how one designs a body for
an animal which is both, say, oviparous, warm-blooded, water-dwelling, four-footed, etc.31
And even with a good solution on our drawing board, this is not always nature´s way. For
sometimes the character or the quantity of matter will dictate another solution. To say that
nature works for the best is here to admit that she will find the best solution all things
considered. But all things are many and variable and the greater the complexity the more
likely it is that the explanation will refer to the indivisible species.
A first reading of T4 might have suggested that descending to the indivisible species
would be the exception. However, we have seen that a complex of differentiae and material
constraints determine the composition of the parts. From this point of view, it is unsurprising
if many of Aristotle´s accounts of the parts will refer to indivisible species. In a nutshell,
Aristotle´s explanatory holism favours accounting for animal parts at the lowest level of
kinds. However, it is important not to overstate this conclusion for at least two reasons.
First, that greater complexity of explanatory factors exerts a downwards pressure
towards explaining the part at the level of the indivisible species does not mean that the
explanation generally descends all the way down to this level. If the aim is to explain why a
part is as it is in certain animals, there can be good reason to pick the ‘commensurate
universal’ as referred to in APo I.4 at a higher level of generality than the indivisible species.
So several insect species have a sting for the same reason, just as it is true generally of hardskinned animals with dry flesh that they have no epiglottis. Here the explanations are rightly
at a higher level of generality than the indivisible species.
Second, as we are considering parts of animals which are not too distant and whose
parts differ only by degrees, differences in degree between parts that are quite small may not
be tracked by Aristotle´s accounts. Think for example of the finer differences between the
feet or beaks of various bird species, which Aristotle passes by in silence.32 There is little
doubt that Aristotle could have given accounts of these differences given his wide arsenal of
30
The correlation between being live-bearing and being warm allows us here to highlight the explanatory factor
amongst a group of differentiae. But this is not always so. There is, for example, a correlation between the shape
of the spleen and the form of an animal´s feet, but no indication of why that should be so. ‘The horn-bearing and
cloven-hoofed animals, such as goat, sheep, and each of the others, have a round spleen, unless, on account of
the animal's size, it grows more quickly lengthwise, as happens with the spleen of the ox. The many-toed
animals, however, such as pig, human being, and dog, all have a long spleen, while those with solid hoofs, e.g.
horse, mule, and ass, have a spleen intermediate between these and a mixture of the two; compared with the one
group they have a wide one, but compared with the other a narrow one.’ (III.12, 673b32-674a4)
31
The differentiae here may be variously ordered according to their importance to the part to be explained. As I
have been at pains to emphasize in this paper the differentiae are not necessarily part of the essence of the
animal though they may follow in various ways from the essence. We should not expect therefore that the order
of the differentiae in the definition of the animal kind is followed in the explanation of a particular part of that
animal´s body.
32
Beautifully illustrated in Lennox 2001c: 163-4.
16
explanatory terms, yet the differences of degree may be too slight to merit attention since our
focus is not on the features of particular species as such. So long or short beak, straight or
crooked, is of interest (662b5-16), but the smaller variations of short or crooked beaks that
characterize particular species fall under the radar. Again, we should bear in mind that
Aristotle focuses on explaining the differences between the parts that obtain between animals,
rather than the differences between species as such. The descent to the indivisible species,
when it occurs, is motivated by the need to explain significant variation in the parts.
To sum up. Aristotle in the PA is interested in explaining the body parts of animals.
This requires direct or indirect reference to essential features of animals, to which those parts
contribute, directly or directly. Some, like blood or flesh, can be directly explained by
reference to the form and function of general animal kinds, nutrition or touch respectively.
Other parts are explained in relation to these parts, bones or sinews, for example, which in
turn may serve to explain further features within the body. Differences between the parts in
different kinds of animal can sometimes be explained at high level of generality. Here
differentia invoked may be one or few, like the difference between oviparous or viviparous,
blooded or non-blooded. However, sometimes the explanation calls for a wider range of
explanatory factors spanning function, life form, environment, matter, etc. The greater the
complexity of explanatory factors required, the more the explanation tends towards referring
to the indivisible species. For it is at this level that most of the different differentiae combine
since it is at this level that all the different constraints on the realization of form and function
ultimately have to be reconciled. The tendency towards the indivisible species is in this
respect driven by considerations of explanatory adequacy. But it is not hard to discern behind
this explanatory holism, Aristotle´s insistence against Plato that what is really there is the
fully differentiated, indivisible enmattered species.
17
Literature cited:
D.Balme (1987), ‘Aristotle’ Biology Was Not Essentialist’, A,Gotthelf and J.Lennox (eds.),
Philosophical Issues in Aristotle’s Biology, CUP, 291-312.
A.Gotthelf (1997), ‘The Elephant’s Nose: Further Reflections on the Axiomatic Structure of
Biological Explanation in Aristotle’ in W. Kullmann and S. Follinger (eds.) Aristotelische
Biologie. Steiner.
A.Gotthelf (2012), Teleology, First Principles and Scientific Method in Aristotle´s Biology,
OUP.
D.Henry, ‘Parts of Animals Book 1 on Methods of Inquiry’ in S.Connell (ed.), The
Cambridge Companion to Aristotle’s Biology, CUP 2021, 83-96.
T.K.Johansen (1997), Aristotle on the Sense-Organs, CUP.
T.K.Johansen (2012), The Powers of Aristotle´s Soul, OUP.
J.Lennox (1997), ‘Nature does nothing in vain…’ in H. Gunther and A. Rengakos Beiträge
zur antiken Philosophie: Festschrift für Wolfgang Kullmann. Steiner, 199-214.
J.Lennox (1999), ‘The Place of Mankind in Aristotle´s Zoology’, Philosophical Topics 27, 116.
J.Lennox (2001a), Aristotle. On the Parts of Animals, OUP.
J.Lennox (2001b) ,‘Divide and Explain: The Posterior Analytics in Practice’, in J.Lennox,
Aristotle’s Philosophy of Biology, CUP.
J.Lennox (2001c), ‘Kinds, Forms of Kinds, and the More and the Less in Aristotle´s
Biology’, in J.Lennox, Aristotle’s Philosophy of Biology, CUP, 160-181.
J.Lennox, (2010), ‘Bios and explanatory unity in Aristotle’s biology’ in D.Charles (ed.)
Definition in Greek Philosophy, OUP, 329-55.
G.E.R.Lloyd (1996), Aristotelian Explorations, CUP.
18
Thomas Kjeller Johansen