Aristotle

Robin Waterfield and David Bostock (eds), Oxford World's Classics: Aristotle: Physics

Main Text

pg 118V

CHANGE

1. Coincidental changes

224a21One kind of change is coincidental change* (as, for example,

when we say that it is an educated person who is walking,

because what is walking is by coincidence someone educated).

23Secondly, we do say that something is changing tout court

when some aspect of it is changing, as when we describe

things as changing because their parts are changing (for ex-

ample, health is restored to the body because it is restored to

26the eye or the chest—i.e. to parts of the whole body). Thirdly,

however, there is the kind of change where something is not

changing coincidentally or because something else, one of its

parts, is changing, but because it itself is immediately chang-

ing. Such a thing is something which is in its own right

capable of change, but it is differently described depending

on the kind of change it is capable of. For instance, it might

be capable of alteration, and more precisely of becoming

healthy or hot.

30The same goes for the agent of change as well. It can impart

change either coincidentally, or by a part (when some aspect

of it is causing the change), or immediately and in its own

right (as, for instance, when a doctor heals or a hand hits).

34There is, then, an immediate agent of change and an object

which is being changed. Then there is the time when the

change is happening, and since every change is from some-

thing and to something else, we also have to take into con-

sideration the starting-point and the end-point of the change.

224b1The object which is immediately changing, the starting-point

of the change and its end-point are all different. Consider

wood, heat, and cold; they might be respectively the object,

the end-point, and the starting-point. The change obviously

takes place in the wood, not in its form, because form (like

place and quantity) does not cause change* and is not changed

pg 119either. No, there is that which causes change, that which is

changed, and the end-point of the change. I say this because

a change is described by its end-point rather than its starting-

point. That is why destruction is a change to a state of non-

existence, despite the fact that anything which is being

destroyed is changing from a state of existence, and genera-

tion is change to a state of existence, despite the fact that it

is from a state of non-existence as well.

10I have already* stated what change is. Now, the end-points

(which may be forms or affections or place*) are not subject

to change. How could knowledge or heat, for instance, be

subject to change? (All the same, one might wonder whether

affections* might not be processes of change, and whether

pallor might not be such an affection. If so, we would have

a case of changing to a change. But presumably it is paling,

16not pallor, which is a process of change.) Again, these end-

points may be the end-points of change either coincidentally,

or because of a part (i.e. depending on something other than

just themselves), or immediately (i.e. not depending on some-

thing other than themselves). For instance, something which

is becoming white might coincidentally change to an object of

thought* (since by coincidence its colour is being thought

about), or it might change to a colour in the sense that white

is a part of colour (and Europe might be the end-point of a

change because Athens is a part of Europe), or it might in its

own right change to the colour white.

22It is clear, then, what it means to say that something is

changing or causing change in its own right, what it means

to say that it is changing or causing change coincidentally,

and what the difference is between its changing or causing

change thanks to something other than itself and its changing

or causing change immediately, because of itself. It is also

clear that change does not take place in the form, but in that

which is being changed—which is to say, something capable

of change when it is actually changing.

26Now, I propose to ignore coincidental change, because it is

a constant aspect of everything in every respect. However, non-

coincidental change is restricted to things that are opposites

or intermediates, or in contradiction.* A survey of examples

pg 120would convince one of this. Things can change from an inter-

mediate state because it acts as the opposite of either of the

two extremes, since it is in a sense the extremes. This also

explains why we talk as if there were opposition between an

intermediate and the extremes, and between the extremes and

an intermediate; for example, the middle string makes a high

note relative to the lowest string, and a low note relative to

the highest string, and grey is pale relative to black and dark

relative to white.

The distinction between change and variation

35All change is from something to something. The word itself*

shows this: one thing comes 'after' another thing—that is, there

is an earlier phase and a later phase. Since all change is from

something to something, there are four possible ways in which

it might occur. There is either change from an entity to an

entity, or from an entity to a non-entity, or from a non-entity

to an entity, or from a non-entity to a non-entity. By 'an entity'

I mean something signified by an affirmative term. It neces-

sarily follows from this that there are three kinds of change:

from an entity to an entity, from an entity to a non-entity,

and from a non-entity to an entity. Change from a non-entity

to a non-entity is impossible* because there is no opposition

involved: they are neither opposites nor contradictories.

225a12Change from a non-entity to an entity, where contradiction

is involved, is coming to be; it is coming to be in an unqualified

sense when the change is unqualified, and it is coming to be

of a particular kind when the change is of a particular kind.

For example, when something is changed from not being pale

to being pale, this is the coming to be of this particular quality,

whereas when something is changed from simply not being to

being a substance, this is simple coming to be; hence we say

that it simply comes to be, not that it comes to be something.

17Change from an entity to a non-entity is ceasing to be; it

is simple ceasing to be when it is change from being a substance

to not being, and a particular kind of ceasing to be when it

is change to an opposite negation, as I have already explained

in the case of coming to be.

pg 12120'Not being' is ambiguous.* Something which 'is not' in the

sense that it is a false combination or separation of subject

and predicate cannot vary, and nor can anything which 'is

not' in the sense that it is only potentially (i.e. in the sense

that it is opposed to that which actually is, tout court). For

although something not pale or not good can still vary coin-

cidentally (since 'something not pale' might be a person),

nevertheless something which just is not an individual thing

cannot vary at all. It is impossible, then, for that which does

26not exist to vary. It follows from this that it is also impossible

for coming to be to be a kind of variation, because it is some-

thing that does not exist that comes to be. I mean, however

true it may be that it is coincidentally coming to be,* it still

remains true to say that not being does belong to that which

simply comes to be. And by the same token, that which does

not exist cannot be at rest either.* Apart from these awkward

consequences, everything that undergoes variation is in some

place or other, but something non-existent is not in place,

because then it would be somewhere. So ceasing to be is not

a kind of variation either, because the opposite of a variation

is either a variation or a state of rest, but ceasing to be is the

opposite of coming to be.

34Since every variation is a kind of change, and there are

three kinds of change (as already mentioned), and of these

the ones involving coming to be and ceasing to be are not

variations—these are the ones which involve contradiction—

it necessarily follows that the kind of change which is from

an entity to an entity is the only one that is a variation. And

entities are either opposites or intermediates (assuming that

we may take a privation* as an opposite) and are signified by

affirmative terms such as 'naked', 'toothless', and 'dark'.

2. The different kinds of variation*

225b5Now, predications are divided into the various categories—

that is, into predications of substance, quality, place, relation,

quantity, and action or affection.* So it necessarily follows

that there are three kinds of variation—qualitative variation,

10quantitative variation, and variation of place. There is no

pg 12211variation in respect of substance, because nothing that exists

is opposite to a substance. Nor is there any variation in re-

spect of relation, because when one of two things which are

relative to each other changes, it is possible for the relation

to cease to obtain^† even though the other thing is not chang-

ing at all, with the result that their variation is coincidental.

Nor is there variation in respect of the action of what is

acting (i.e. the agent of variation) or the affection of what is

acted on (i.e. the thing that varies), because a variation can-

not vary and a coming to be cannot come to be; to put it

generally, a change cannot be changed.

16The first point to notice here is that there are two ways in

which variation of a variation might be possible. The variation

which is varying might be an underlying thing; this is how a

person varies when he changes to being dark instead of pale.

But can a variation too become hot or cold in this way, or

alter its place, or increase or decrease? No, this is impossible,

21because change is not an underlying thing. Alternatively, the

variation might vary because some underlying thing (which is

not to be identified with the variation itself) changes from a

process of change to a different species of existence. But this

is impossible too, except coincidentally, because variation itself

is change from one form to another, as when a person changes

from illness to health. (The same goes for coming to be and

ceasing to be as well, except that they involve opposites of one

27kind, whereas variation involves opposites of another kind.) It

follows that the person of our example, at the same time as

changing from health to illness, would also be changing from

this particular change to some other kind of change. So ob-

viously, by the time he is ill, he will have changed to what-

ever other change it may be (it could even be a state of rest),

and moreover the change he changes to will not be just any

kind of change;^† for that too—the process of changing from

one change to another—must be a change from something to

something opposite, and so what he changes to will be the op-

posite process—namely, becoming healthy.* No, a change can

only change coincidentally, as when there is a change from re-

membering to forgetting* because the subject involved changes

at one time to knowing and at another time to ignorance.

pg 12333Secondly, the idea that a change could change and a coming

to be could come to be generates an infinite regress. If the later

change is to happen, then the one that comes before it must

happen too. For instance, if a simple coming to be was itself

coming to be at some time, then the thing that would come to

be was also becoming a thing that comes to be, and so there

was not yet anything which was simply coming to be, but

only a thing becoming something (i.e. becoming a thing that

comes to be). And if this too was also coming to be at some

time, then at that time it was not yet even becoming a thing

that comes to be. And since there can be no first term in an

infinite series, then this sequence cannot start anywhere, and

so it cannot continue either. Consequently, on this hypothesis

coming to be, variation, and change are completely impossible.

226a6Thirdly,* a single changing thing also has the capacity for

the opposite kind of change (and also for the state of rest

which is opposite to its change); and similarly coming to be

and ceasing to be are properties of the same thing. Now,

anything which is coming to be a thing that comes to be

cannot be ceasing to be that thing just as it is coming to be

it, because in order for anything to cease to be, it first has to

be, and it cannot do so after it has come to be either; therefore

it is ceasing to be a thing that comes to be at the very time

that it has become it!

10Fourthly,* anything which is coming to be and anything

which is changing have to have an underlying matter. On the

present hypothesis, what will this be? Just as it is either a

body or a mind which is capable of alteration, what will it

be that becomes a change or becomes a coming to be? And

again, what will it be that they change to? After all, the

change or the coming to be of any specific thing has to be a

change from one state to another. Moreover, what kind of

changes will these be? The coming to be of learning cannot

be learning, and so also the coming to be of coming to be

cannot be coming to be, and the coming to be of anything

cannot be that thing either. Also, both the underlying thing

and the end-point of the change must be one of our three

forms of variation, and so, for instance, movement must itself

alter or move.

pg 12419In short, then, since everything that changes changes in one

of three ways—either coincidentally or by a part changing or

in its own right—the only way in which change can change

is coincidentally (as when someone who is recovering from

illness runs or learns); but we dismissed coincidental change

a long time ago.

23Since there can be no variation of substance or relation or

action and affection, it is only in respect of quality and quan-

tity and place that there can be variation. For each of these

26categories admits opposition.* Let variation of quality be 'al-

teration', since this is a general term which links both the

opposites together. By 'quality' I do not mean here the kind

of quality which is an essential property of substance (for the

differentia of a substance is a quality), but an affective quality,*

which allows us to describe something as affected or as incap-

29able of being affected. There is no general term for variation

of quantity, but it is described by reference to each opposite

separately as 'increase' or as 'decrease', increase being change

towards the completion of a thing's size and decrease being

32change away from this. As for variation of place, there is no

general term which covers both the opposites together or each

of them separately, but let us use 'movement' as the general

term, despite the fact that,* strictly speaking, only those things

are said to be moving which are such that, once they are

changing place, they do not have the power to stop by them-

selves, and which do not initiate their own change of place.

226b1Change within a single form—that is, change to a different

degree of that form—is alteration, since alteration is, in a

qualified or unqualified sense, change from one opposite or to

the other. When the direction of the change is towards a lesser

degree of the form in question, we tend to say that the end-

point of the change is the opposite of that form, and when

the direction of the change is towards a greater degree, we tend

to say that the starting-point of the change is the opposite of

the form and that the end-point is the form itself. The point

is that it does not make any difference whether the change is

qualified or unqualified, except that in the qualified version

the opposites will have to be present in a qualified sense. The

form is present to a greater or lesser degree, depending on

pg 125whether or not there is more or less of the opposite present

in it.

8These arguments show that there are only the three kinds

of change. Now, something is 'unchanging' not only if it is

completely incapable of change (compare the invisibility of

sound), but also if it takes a lot of time and effort to get it

changing or if its change is initially sluggish (when it is de-

scribed as 'hard to change'). Then there is also that which is

by its nature capable of change, but which is not changing

when, where, and how it is its nature to change. This is the

only unchanging thing I describe as being at rest, since rest is

the opposite of change and so will be the privation of change

in that which is capable of admitting change.

16I have now explained what change is, what rest is, how

many kinds of change there are, and what kinds of change

there are.

3. Definitions of various terms

18Next, we had better define 'together' and 'apart', 'in contact',

'between', 'successive', 'consecutive', and 'continuous', and

explain the kinds of situation to which each of them is by its

21nature applicable. I say that things are together in respect of

place when they coincide in a single immediate place,* apart

when they are in different immediate places, and in contact

when their extremes are together.

227a7Now, every change involves opposites, and opposites can

be either contraries or contradictories; but since there is noth-

ing intermediate between contradictories, it obviously follows

that there must be contraries for there to be something in

between. So between involves at least three terms. For the

final stage of a change is the opposite, and that which is in

between is what the changing thing naturally reaches before

it reaches the final stage of its changing, assuming that the

process of change is continuous and in accordance with the

226b27thing's nature. The change is continuous* if none of the pro-

cess is left out, or only a very little. It is what happens to the

process in which the change occurs, not to the time, which is

relevant, because there is nothing to prevent some time being

pg 126left out, and moreover it is perfectly possible to sound the

31highest note immediately after the lowest note. Change of

place makes clear what I mean, but it is equally evident in

other kinds of change as well. In change of place, 'opposite'

refers to that which is furthest away in a straight line, be-

cause a straight line is the shortest distance between two

points and is therefore limited; so it acts as a measure, just

as anything limited does.

34A thing is successive when it comes after a beginning, either

in position or in form or by some other distinguishing criterion,

and when there is nothing of the same kind as itself between

it and that to which it is successive. Think of a line (or a

number of lines) succeeding a line, or a unit (or a number of

units) succeeding a unit, or a house succeeding a house: it

does not matter if something of a different nature comes in

between. The point is that for anything to be successive, it

has to succeed something and it has to come later than that

thing. After all, the number one is not successive to the number

two, and first day of a month is not successive to the second

day either; it is the other way round.

227a6A thing is consecutive if it is both successive and in contact.

Something continuous is consecutive in a sense, but I say that

something is continuous rather than consecutive when the

limits by which the two objects are in contact have become

identical and, as the word implies, enable one object to con-

tinue into the other. This is impossible where there are two

separate limits. It is clear from this definition that continuity

is a property of things which naturally form a unity by their

contact with one another. And the way in which what makes

them continuous* is single determines the way in which the

whole is single too; this may involve pinning, for instance, or

gluing, or contact, or grafting.

17It is also clear that successiveness is primary. For contact

inevitably implies successiveness,* but successiveness does not

necessarily imply contact. (That is why successiveness is a

property of things which are prior by definition—numbers,

for instance—but contact is not.) Also, wherever there is

continuity there is contact, but not vice versa, because the

fact that the extremes of the things involved are together does

pg 127not necessarily mean that they are one. If they are one, how-

ever, they are bound to be together as well. And so growing

together is the last of the series, because any extremes that

grow together are bound to be in contact, but the fact that

things are in contact does not necessarily mean that they are

growing together. Where there is no contact, however, there

27is obviously no growing together either. It follows from this

that even if, as some say, a point and a unit both have inde-

pendent existence, they cannot be identified, because contact

is a property of points,* whereas successiveness is a property

of units, and also because there can be something between

points (every line is between points), whereas there cannot be

anything between units (nothing comes between one and two).

32I have now defined 'together' and 'apart', 'contact', 'between',

'successive', 'consecutive' and 'continuous', and I have explained

the kinds of situation to which each of them is applicable.

4. What counts as a single change

227b3To say that a change is one is ambiguous, because 'one' is

ambiguous. Generic unity of change depends on the categories

of predication; any movement is generically the same as any

other movement, but alteration is generically different from

movement. A change is specifically the same as another change,

however, when it is not only generically identical, but also is

a change in the same indivisible species. For example, there

are specific differences within the genus colour, and that is

why becoming black is specifically different from becoming

white, but there are no specific differences within whiteness.

So any instance of becoming white is specifically the same as

any other.

11(If there are things which are species as well as genera,

there is obviously a sense in which any change which falls

within such a genus is specifically identical to any other, but

not unqualifiedly so. Consider learning, for example, given

that knowledge is a species of apprehension as well as a

genus consisting of the various branches of knowledge.)

14Someone might wonder whether what it takes for the change

to be specifically one is that the same thing changes from the

pg 128same starting-point to the same end-point (think, for instance,

of a single point moving from this place to that over and over

again). However, if this is so, circular movement will be the

same as movement in a straight line and rolling will be the

same as walking. Or have we already decided* that where

what the change is in is specifically different, the change is

specifically different, and that a circular route is specifically

different from a straight route?

20That is what it is for change to be generically and spe-

cifically one, and it is simply one, without any qualification,

when it is one in definition and in number. The following

distinctions will show what kind of change this is. There are

three factors we take into account when talking about

change—what, in what respect, and when. What I mean is

that there has to be something which is changing (a person,

say, or gold), and it has to be changing in a certain respect

(in place, perhaps, or in some quality), and it has to be chang-

ing at a certain time, because every change happens in time.

27Of these, the respect is responsible for the generic or specific

unity of the change, and time is responsible for the consecu-

tiveness of the change; but all of them together are respons-

ible for unqualified unity. For this, the respect (the species of

the change) must be one and indivisible, the when (the time

of the change) has to be single with nothing left out, and the

changing thing has to be single, and not coincidentally so.

Think of a pale object turning dark, and of Coriscus walking:

the pale thing may be the same thing as Coriscus, but that is

coincidental. Moreover the changing thing must be single,

not just in the sense that a single common character is in-

volved; two people might simultaneously be recovering from

the same ailment (an eye infection, say), but this process of

recovery is not absolutely single, only specifically so.

228a3But what if Socrates alters, with an alteration which is

specifically the same, and does so first at one time and then

again at another time? If it is possible for something which

has ceased to be to come into being again as numerically the

same one thing, then this alteration of Socrates' could be one

as well; if it is impossible, however, it is the same alteration,

but not a single alteration.*

pg 1296There is a problem which is closely related to this one,

namely whether health and bodily states and affections in

general are one in substance.* The problem arises because

their seats are evidently changing and in flux. If the health I

had in the morning and the health I have now are one and

the same, why should health which is restored after an inter-

val not be the same health too? Why should that health and

this health not be numerically one? After all, the argument is

12the same in both cases. There is this much difference between

them,* however: if there are two states, because they are the

states of individuals which are in this way numerically two,^†

there must be two activities as well, since an activity is nu-

merically one only if it is the activity of something numeric-

ally one. But if there is one state, that still might not lead us

to count the activity as one as well, because when a person

stops walking, the walking no longer exists, but it will exist

again if he walks again. So if the walking is one and the

same, it would have to be possible for something which is

one and the same to cease to exist and then to exist again

time after time. But these problems lie outside our present

enquiry.

20Every change is continuous,* since every change is divisible,

and so change which is unqualifiedly one must be continuous

too, and if it is continuous, it must be one. For instance, it

is not the case that every change is continuous with every

other change, just as it is not the case that any two objects

taken at random are continuous; two things are continuous

only if their limits are one. Now, some things do not have

limits, and although other things do, their limits are different

in form from one another and share only a name. How could

the end of a line and the end of a walk be in contact or

26become one? Changes which are specifically or generically

different could be consecutive to one another: someone could

go for a run and then immediately come down with a fever.

And there could be movements which are consecutive, such

as the race where torches are passed on in relay. They are not

continuous, however. After all, we have established that con-

tinuity involves unity of limits. So continuity of time makes

changes consecutive and successive, but it is continuity of the

pg 130changes themselves that makes them continuous—that is, when

228b1their two limits are the same. That is why for a change to be

continuous and one in an unqualified sense it has to be speci-

fically the same, there has to be only one changing thing

involved, and the change has to happen in a single period of

time. It has to happen in a single period of time, because other-

wise there will be a period of non-change in between, since

when change is interrupted, there is bound to be rest. Any

change which is interrupted by rest is a number of changes,

not a single change, and so a change which involves intervals

of rest is neither single nor continuous, and these intervals

7occur if there are intervening periods of time. But even if the

time involved is continuous, as long as the change is not speci-

fically the same, it makes no difference that only a single stretch

of time is involved: the change remains specifically different.

The point is that although specific identity is a prerequisite

for a change to be single, it is not the case that a change that

is specifically single is necessarily unqualifiedly one.

11I have now explained what it is for a change to be un-

qualifiedly one. Moreover, a change is also described as one—

generically, specifically, or in substance*—if it is complete,

just as in other cases completeness and wholeness are properties

of something that is one. But sometimes a change is described

as one even if it is incomplete, as long as it is continuous.

15A uniform change is also described as one, though not in

the same way as the changes we have already discussed. What

I mean is that there is a sense in which a non-uniform change

gives the impression of not being single, whereas this descrip-

tion seems more suited to a uniform change (as it is to a

straight line, for instance). After all, a non-uniform change is

divisible. But non-uniform changes seem to differ in degree,

some being more uniform than others. Every kind of change

can be either uniform or non-uniform. An object can alter uni-

formly, it can move uniformly (around a circle, for instance, or

in a straight line), and the same goes for increase and decrease.

Non-uniformity is an inconstancy which may occur in the

path the change is taking, since it is impossible for a change

to be uniform if it is moving over a non-uniform magnitude,

such as an angled line or a spiral or any other magnitude

pg 131which is not such that any two parts, chosen at random, will

25fit on to one another. Alternatively, non-uniformity may occur

not where the change takes place,^† nor in the time, nor in the

end-point, but in the way the change happens. For instance,

there may be inconstancy of speed; any change which happens

at the same speed is uniform, whereas any change where the

speed differs is non-uniform. That is why quickness and slow-

ness are neither species nor differentiae of change: it is because

they are found in all the various species of change. It follows*

that greater or lesser heaviness and lightness, considered as

tendencies in a single direction (i.e. when one compares the

relative weight of one piece of earth with that of another, or

the relative weight of two bits of fire), do not constitute

different species of change either.

229a1Although non-uniform change is single because it is con-

tinuous, then, it is less single than uniform change, as we can

see in the case of angled movement; and 'less' always implies

the infiltration of the opposite. Now, every change that is

single can be either uniform or non-uniform, and therefore*

changes which are not specifically the same but are consec-

utive with one another cannot form a single, continuous unity.

How could a change which is a combination of alteration

and movement be uniform? In order to be uniform, its parts

would have to fit on to one another.

5. How a change is opposite to a change

7We should also decide which changes are opposite to which,

and do the same for rest as well. First, what is it for one

change to be the opposite of another? Is it that the starting-

point of one is the end-point of the other (one might be a

change from health, for instance, while the other is a change

to health)? This is what seems to apply to coming to be and

ceasing to be. Or is it that they have opposite starting-points

(one being a change from health, say, while the other is from

illness)? Or is it that they have opposite end-points (one being

a change to health, for instance, while the other is a change

to illness)? Or is it that the starting-point of one is the opposite

of the end-point of the other (one being a change from health,

pg 132for instance, while the other is a change to illness)? Or is it

that both their end-points and starting-points are opposed to

one another (one being a change from health to illness, for

instance, while the other is a change from illness to health)?

One or more of these kinds of opposition must apply, since

the list is exhaustive.

16Now, changes where the starting-point of one is the opposite

of the end-point of the other (where one, for instance, is a

change from health, while the other is a change to illness) are

not opposites: they are identical, in fact (except that what it

is to be one is different from what it is to be the other, in the

sense that it is not the same to change from health and to

change to illness).

20Also, changes with opposite starting-points are not opposites

either, since a change from an opposite is at the same time a

change to the opposite of that opposite (or to something in

between the two opposites)—but I will discuss this in a

moment.*

22Opposite end-points would seem to account for opposition

between changes better than opposite starting-points, since in

the latter case there is a loss of opposition, whereas in the

former case opposition is gained. Besides, we do describe any

given change in terms of its end-point rather than its starting-

point: the change to health, for instance, is called 'recovering

health', and the change to illness is called 'falling ill'.

27We are left, then, with changes to opposite end-points, and

those with both opposite end-points and opposite starting-

points. Now, it may be that changes to opposite end-points

are also changes from opposite starting-points, although what

it is to be the one is perhaps different from what it is to be the

other. What I mean is that the change to health is different

from the change from illness, and the change from health is

different from the change to illness. But since change and vari-

ation are different (in the sense that variation is change from

an entity to an entity), then variations whose end-points and

starting-points are both opposed are opposites; for example,

the variation from health to illness is the opposite of the

229b2variation from illness to health. A survey of particular examples

shows the kinds of cases which are taken to be opposites. For

pg 133instance, falling ill is taken to be the opposite of recovering

health, and being taught the facts is taken to be the opposite

of being misled by someone else, since the end-points are

opposed (after all, just as one can acquire the truth from

someone else, as well as from oneself, so one can also be

misled by someone else or by oneself). Or again, movement

upwards is taken to be the opposite of movement downwards

(since the end-points are opposed on the dimension of length),

movement to the right is taken to be the opposite of move-

ment to the left (since the end-points are opposed on the

dimension of breadth), and movement to the front is taken to

be the opposite of movement to the back (since here too there

are opposite end-points).

10However, processes in which only the end-point is an op-

posite (as when pallor comes into being, but not from any-

thing) are changes, but not variations.* And in cases where

there is no opposite, one change is the opposite of the other

if its starting-point is the end-point of the other; that is why

coming to be is the opposite of ceasing to be, and loss is the

14opposite of gain. But these are changes, not variations. In

cases where the opposites have an intermediate, variations to

the intermediate point should be counted as variations to

opposites, in a sense, because as far as the variation is con-

cerned, the intermediate point acts as an opposite, in which-

ever direction the change is proceeding. For instance, grey

acts as black in the change from grey to white, and as black

again in the change from white to grey, but it acts as white

in the change from black to grey. For, as I have already said,

compared to either of the two extremes, the mid-point is

described as in a sense being either of them.

21So the way in which changes are opposed to one another

is if both their starting-points and their end-points are op-

posed to one another.

6. How change and rest are opposites

23However, change is not only opposed by change: rest seems

to be the opposite of change too. So here is another issue we

had better settle. A change is opposed in an unqualified sense

pg 134by a change, but rest is also opposed to it in that it is the

privation of change, and there is a sense in which we describe

the privation of anything as its opposite. Now, the opposite

of change of a particular kind is rest of the same kind; for

instance, the opposite of change of place is staying in one

place. But now this statement needs qualifying. Which is the

opposite of staying in a given place? Is it movement to that

place or movement from that place? Clearly (remembering

that two entities* are involved in movement), staying in place

A is opposed by moving from there to place B, and staying

in place B is opposed by moving from there to place A.

31At the same time, staying in place A and staying in place

B are also the opposites of each other. After all, there could

hardly be opposition between changes and not between states

of rest; and rest in an opposite state is an opposite state of

rest. For instance, remaining in a state of health is the opposite

of remaining in a state of illness (as well as being the opposite

of the change from health to illness. It would be ridiculous

for it to be opposed by the change from illness to health,

because the change whose end-point is the state of remaining

healthy is a process of coming to rest rather than the opposite

of the rest, or rather the process of coming to rest happens

to coincide with the change. But it must be opposed by either

the change from health to illness or the change from illness

to health.) After all, it cannot be remaining in the state of

pallor that is the opposite of remaining in the state of health.

230a7In cases where there is no opposite, the change where some-

thing is the starting-point is the opposite of the change where

that thing is the end-point; this is not a variation, however.

Change from being, for instance, is the opposite of change to

being. Also, in these cases there is no such thing as rest,* but

only changelessness. And if some entity was involved, its

changelessness in existence would be the opposite of its change-

12lessness in non-existence. But one might object that there is

no such thing as a non-existent thing, and ask what change-

lessness in existence is opposed to, and whether changelessness

in existence is in fact a state of rest. But if it is, then either

it is possible for the opposite of a state of rest not to be a

variation, or else coming to be and ceasing to be are variations.

pg 135Clearly, then, since coming to be and ceasing to be are not

in fact variations, we should not describe changelessness in

existence as a state of rest; we should acknowledge its simi-

larity to a state of rest, but call it a state of changelessness.

And it will be the opposite either of nothing or of changeless-

ness in non-existence or of ceasing to exist, because ceasing

to exist is a change which has this state of changelessness as

its starting-point, while coming to be is a change which has

it as its end-point.

The opposition between natural and unnatural change and rest

18Why, someone might ask, should both rest and change be

either natural or unnatural in the case of change of place,

while this does not obtain for other kinds of change? For

instance, the contrast between natural and unnatural does

not apply to alterations: recovering health is no more natural

or unnatural than falling ill, and paling is no more natural or

23unnatural than darkening. The same goes for increase and

decrease: they are not opposed to each other in the sense that

while one is natural, the other is unnatural, and one increase

is not opposed to another increase in that way either. The

same argument applies to coming to be and ceasing to be: it

is not the case that coming to be is natural, while ceasing to

be is unnatural (after all, growing old is natural), and it is

also evident that the contrast between natural and unnatural

does not apply to instances of coming to be.

29However, if what happens by force is unnatural, then forced

ceasing to be is unnatural, and is opposed to natural ceasing

to be. There are also some instances of coming to be that are

forced and are not as destined (and are therefore opposed to

natural instances). And are there not forced instances of in-

crease and decrease, such as the rapid growth to maturity of

those who have been given a luxurious diet, or the ripening^†

of grain even when it is not compressed?* And what about

alteration? Is it not the same for it too? If so, some alterations

would be forced, while others are natural. An example of the

difference between the two kinds of alteration might be people

recovering from fever on non-critical days, as distinct from

pg 136those who do so on the critical days: the second kind of

alteration would be natural, the first unnatural.

230b6So there will also be instances of ceasing to be which are

opposed to one another, not to instances of coming to be.

Why not? Why should there not be a sense in which this is

so? After all, one instance of ceasing to be may be enjoyable,

while another is distressing. In other words, it is not the case

that instances of ceasing to be are opposed to each other in

an unqualified sense, but in the sense that one of them has

one quality, while the other has a different quality.

10Generally speaking, then, changes are opposed to changes,

and states of rest to states of rest, in the way I have described.

For example, since up and down are opposites in the category

of place, upward movement is the opposite of downward

movement. Fire naturally moves upwards, and this is the op-

posite of the natural movement of earth, which is downwards;

the natural upward movement of fire is the opposite of its

15unnatural downward movement. The same goes for states of

rest: rest high up (unnatural for earth) is the opposite of

movement down from high up (which is natural for earth); so

unnatural rest is the opposite of a thing's natural movement.

After all, a thing's movements are opposed on the same prin-

ciple: one of its movements—either upward or downward—

is natural, while the other is unnatural.

21There is a problem here: if a state of rest is not everlasting,

does it come into being and is its coming into being the same

as coming to rest? Coming into being would then be some-

thing that happened to unnatural states of rest (such as earth

being at rest high up), and it would follow that when earth

is forced to move upwards, it is coming to rest. But every

process of coming to rest seems to involve the object accel-

erating, whereas the opposite is the case when a thing has

been forced to move. So it would be at a standstill without

having come to a standstill. Also, a thing's coming to rest is

generally taken to be either the same as its moving to its

proper place, or at least to be something that coincidentally

happens at the same time.

28It might be objected that rest in a given place might not be

the opposite of movement from that place, on the grounds

pg 137that when something is moving away from a place and losing

touch with it, it seems to retain what is being left behind;*

therefore, it might be said, if the state of rest in question is

the opposite of movement from there to somewhere which is

opposed to there, the object will have two opposites simultan-

eously. However, surely it is in some sense at rest, provided

that it remains in its former state; and, to put the matter

generally, whenever something is changing, part of it is at the

starting-point of the change and part of it is going towards

the end-point. That is also why the opposite of a change is

better regarded as another change, rather than a state of rest.

231a2I have now explained in what sense change and rest are

each single, and what kinds of opposition exist among them.

Notes Settings

Editor’s Note

224^a21 coincidental change: as with causes, so also with changes, Aristotle holds that there is some one preferred way of describing the agent of change, the object changed, and the initial and final states of the change. The preferred description is said to describe 'a change in its own right', and other descriptions are said to describe 'a coincidental change'. Again, as with causes, Aristotle gives no criteria for the preferred description, but we get some idea of what he has in mind by attending to his examples.

Editor’s Note

224^b5 form … does not cause change: presumably what Aristotle means is that form is not an agent of change; the Greek word here translated 'cause change' could also be rendered 'set a change in motion'. Even so, this claim is not too easy to reconcile with the claim of Book II that form should be counted as efficient cause and formal cause and (in particular) final cause of many natural changes (198^a24). See also note on 224^b11.

Editor’s Note

224^b10 already: in Chs. 1–2 of Book III.

Editor’s Note

224^b11 forms or affections or place: probably 'form' is here to be understood as limited to 'substantial forms', i.e. to predicates in the category of substance. Also 'affection' is used here for qualities, and quantities are omitted from the list. (Similarly at ^b5 above, where qualities are omitted from the list.) But at ^b25 below, 'form' is apparently used to cover all four of the categories in which change takes place, as in Ch. 7 of Book I. (Cf. 225^b24.)

Editor’s Note

2241^b14 affections: as in English, the Greek word 'affection' can be used both for the process of being affected by something and for the product of that process.

Editor’s Note

224^b18 change to an object of thought: suppose that I bleach the shirt so that it becomes white, and meanwhile you are thinking of the colour white. Of course it does not follow that the shirt comes to be thought of (by you), but it may be said to follow that the shirt comes to be something (namely white) that is being thought of (by you).

Editor’s Note

224^b29 opposites or intermediates, or in contradiction: in Ch. 5 of Book I Aristotle argued that all change is between opposites or intermediates, but in Ch. 7 this was subsumed under the more general claim that all change is between a form and its privation. This included generation and destruction, for in such changes the matter was held to persist throughout, now with the form and now with the privation. But here Aristotle will not draw attention to anything that persists through generation and destruction (save possibly at 225^a27), and he will say that this case, and this case only, is covered by 'contradiction' but not by 'opposites or intermediates'. (See also note on 225^b3.)

Editor’s Note

225^a1 the word itself: the word 'change' (metabolē) is a compound word, with the word 'after' (meta) as its first component.

Editor’s Note

225^a10 change from a non-entity to a non-entity is impossible: what changes from weighing less than a pound to weighing more than a pound must also change from weighing neither r pound nor more to weighing; neither 1 pound nor less, but no doubt Aristotle would dismiss the latter description as merely 'coincidental'.

Editor’s Note

225^a20 'Not being' is ambiguous: in Greek 'is not' can be used to mean 'is not the case', or 'does not exist', or (in context) 'is not so-and-so, e.g. pale'. These are Aristotle's three examples. He claims that what 'is nor' in either of the first two ways cannot vary, but his real concern here is just with what does not exist.

Editor’s Note

225^a27 it is coincidentally coming to be: this is an obscure phrase. Perhaps Aristotle means, in accordance with the doctrine of Ch. 7 of Book I, that there is always a kind of coincidence involved when anything comes to be. For example, when a statue comes to be, that is because some bronze has become statue-shaped, and then this bronze is the statue, but only coincidentally. But one could wish that he had been more explicit.

Editor’s Note

225^a29 cannot be at rest either: when the two standard words for change in Greek are distinguished as 'change' (metabolē) and 'variation' (kinēsis), it is the latter that contrasts with 'rest' (ēremia). So, on Aristotle's account, to be at rest is to be capable of varying but not actually varying. Hence he infers that since what does not exist is not capable of varying, it cannot be said to be at rest either (cf. 261^b11–12).

Editor’s Note

225^b3 a privation: of the examples that follow, it would seem that 'naked' and 'toothless' are examples of opposites that are privations, whereas 'dark' is an example of an opposite that is not a privation. Aristotle claims here that words for privations are positive. (The word for 'toothless' does not have a negative prefix or suffix in Greek.) This may be contrasted with his practice in Ch. 7 of Book I, where the privation is always expressed by a negative term, and with the way that he goes on to identify the privation with not being in Chs. 8 and 9 (191^b15–16, 192^a3–6).

Editor’s Note

225^b5 The different kinds of variation: this heading is inserted where the sense demands it. But the MSS mark the beginning of Ch. 2 as occurring two sentences later, at 225^b10.

Editor’s Note

225^b7 action or affection: note that time is omitted from this list. (See Introduction, p. xlvii.) Note also that the phrase 'action or affection' seems here to be treated as the name of a single category.

Critical Apparatus

225^b12 Reading μεταβάλλοντος 〈μή〉‎ (Schwegler).

Critical Apparatus

225^b30 Reading τυχοῦσαν, δεῖ γὰρ κἀκείνην ἔκ τινος εἴς τι ἕτερον, ὥστε‎ with MS H.

Editor’s Note

225^b31 namely, becoming healthy: the argument is that if a person undergoing the change from being healthy to being ill is to change from undergoing that change to undergoing another, then he must change to the opposite change, i.e. the change from being ill to being healthy. Aristotle apparently thinks that this is contradictory, for the first change will lead to his being ill at the same time as the second makes him healthy. But the truth seems to be that we have here a perfectly coherent case of one who is getting an illness, but this process is then reversed (e.g. by a dose of medicine), so he never does go down with the disease but ends up as healthy as at first.

Editor’s Note

225^b32 a change from remembering to forgetting: Aristotle writes as though he supposes that remembering and forgetting are themselves changes (rather than states). This is strange.

Editor’s Note

226^a6 Thirdly: the argument is this. Suppose that from t₀ to t₁, a thing x is coming to be a thing that comes to be, and that from t₁ to t₂ it is coming to be. The premiss is that a process of coming to be must be answered by a corresponding process of ceasing to be, so there must also be a period when x is ceasing to be a thing that comes to be. But when could this be? Not before t₁, for x does not become a thing that comes to be until t₁; not between t₁, and t₂, for at those times x is a thing coming to be, and so cannot also be ceasing to be such a thing; and finally not after t₂, for by then x has ceased to be a thing coming to be, and so cannot be ceasing to be such. (Compare the puzzle at 218^a16–21 on when an instant of time ceases to be.) However, there does not seem to be anything wrong with the reply that x is ceasing to be a thing that is coming to be during the later part of the period from t₁ to t₂.

Editor’s Note

226^a10 Fourthly: by the doctrine of Ch. 7 of Book I, if anything comes into being there must be some matter that is first in one state and then in another, and the matter in its final state constitutes the thing that comes to be. Aristotle asks what this matter could be, and what its final state is, when what comes into being is itself a change or a coming to be. (The plural 'they change to' at ^a13 shows that he thinks that the matter would have to be different in the two cases under discussion, i.e. the coming into being of (i) a change (i.e. variation) and (ii) a coming into being. The final sentence of the paragraph assumes that the matter, i.e. 'the underlying thing', would itself have to be a change (i.e. variation) of some kind. It is not clear how this assumption is justified.)

Editor’s Note

226^a25 each of these categories admits opposition: some qualities have opposites (e.g. pale and dark) and some do not (e.g. round, square, triangular), but any quality can be the end-point of a change. No specific quantity (such as 3 feet long) has an opposite, but again any such quantity can be the end-point of a change. (The 'opposition' that Aristotle sees in this category is that of 'more and less'.) Similarly, no specific place is opposite to any other (but Aristotle is probably thinking of the opposition between up and down).

Editor’s Note

226^a29 an affective quality: affective qualities are defined at Categories 9^a28–10^a10 as a special class of qualities (cf. note on 244^b5), but here Aristotle presumably means the term to include all qualities except those that are essential to the subject in question. (Loss of an essential quality is destruction.)

Editor’s Note

226^a33 despite the fact that: the word here used for 'movement' is more literally 'being carried', and is standardly used of the movements of non-living things. It would not naturally be used of, e.g., walking or running or jumping.

Editor’s Note

226^b22 when they coincide in a single immediate place: that is, when they completely occupy exactly the same place. Aristotle will apply this notion not to three-dimensional bodies but to limits, e.g. points and surfaces, and two limits which touch do occupy exactly the same place where they touch. (But one may observe that Aristotle's account of place in Book IV does not allow such a limit to have a place.)

Editor’s Note

226^b27 change is continuous: what follows is not the definition of continuity promised at the beginning of the chapter; that is coming later at 227^a10, and it will explain what it is for one object to be continuous with another. Nor is it Aristotle's usual account of what makes a change continuous, for that does require that there be no gaps in the time (as e.g. in the next chapter; see note on 228^a20). Rather, this is an ad hoc meaning for 'continuous', adopted simply to clarify the idea of one thing being between others. The thought is that x is between y and z if a natural change from y to z would go through x.

Editor’s Note

227^a16 what makes them continuous: more literally, what holds them together, since the word 'continuous' literally means 'held together' in Greek. (Cf. note on 220^a10.) Aristotle is thinking that if the limits of x and y 'are identical' where they touch, then the whole which has x and y as parts will move as a piece. But (a) I should not have thought that nailing or gluing two objects together would count as making their limits 'identical', and (b) the top half and the bottom half of the water in the jug surely do have limits that 'are one', but they do not move as a piece.

Editor’s Note

227^a18 contact… implies successiveness: this is a mistake, on the existing definitions, for Aristotle has (very naturally) defined 'successive' in such a way that if x succeeds y then x must come after y in some suitable ordering. But contact does not require any ordering from before to after.

Editor’s Note

227^a29 contact is a property of points: Aristotle cannot mean that one point can touch another; he always (and correctly) denies this. I take it that he must mean, somewhat loosely, that points are involved in contact, for one thing will touch another at one or more points.

Editor’s Note

227^b19 have we already decided: the correct answer seems to be 'No', but perhaps Aristotle thinks that at 227^b7 the phrase 'in the same indivisible species' should be taken as covering 'what the change is in' as used here, i.e. the path of the change (straight or circular) and the manner of the change (by walking or by rolling).

Editor’s Note

228^a6 the same alteration, but not a single alteration: that is, the same in species but not in number.

Editor’s Note

228^a8 one in substance: Aristotle usually uses this phrase to mean 'one in definition' (and it is so translated in the other books of the Physics), but here it apparently means 'one in number'. Cf. 228^b13.

Editor’s Note

228^a12 this much difference between them: the text of the rest of this sentence is very uncertain; the reading adopted here is designed to give a reasonable argument, namely this. Suppose one says that my state of health this morning cannot be the same state as my state of health this afternoon, because they are not states of the same thing, on the ground that they are states of the body and the body is 'changing and in flux' (so it is 'in this way'—i.e. because of the flux—'numerically two'). Then it will follow by the same argument that my activity of walking this morning is equally not the same activity as my activity of walking this afternoon. On the other hand, if we insist that the body is still the same body (despite its flux), then we can maintain that there is only one state of health, existing both in the morning and in the afternoon, while still insisting that there are two distinct activities of walking, one in the morning and a different one in the afternoon. This is because the two activities are separated by a temporal gap, during which there is no walking.

It may be noted that Aristotle here equates activities with changes (such as alterations, ^a3–6). But elsewhere he distinguishes them, and counts walking as a change but not an activity. (The locus classicus is Metaphysics Θ‎6, 1048^b18–35.)

Critical Apparatus

228^a13 Reading διὰ τὸ οὕτως τῷ ἀριθμῷ 〈δυοῖν εἶναι〉‎ (Bostock).

Editor’s Note

228^a20 Every change is continuous: elsewhere Aristotle defines a continuous thing as one that is divisible (only) into parts that are themselves further divisible, and hence as something that is infinitely divisible (De Caelo 268^a6–7; cf. Physics vi. 232^b24–5). That appears to be the definition he is relying on here, when he says: 'since every change is divisible [sc. into smaller changes]'. The definition, however, leaves out a point that he relies on in the opening argument of Book VI, and that he evidently regards as crucial, namely that the parts into which a continuous thing is divided must share their limits I suspect that he would welcome a definition such as this. A continuous change occupies a continuous period of time, i.e. one with no gaps; moreover, at every instant during that period the changing object is in a state that is different from its states at suitably nearby instants (i.e. it is not at rest at any instant in the period); and further, its state at one instant of the period differs by as little as you please from its states at suitably nearby instants (i.e. it never 'leaps' instantaneously from one state to a different state).

Editor’s Note

228^b12 generically, specifically, or in substance: see note on 228^a8.

Critical Apparatus

228^b26 Retaining ποῦ‎ with the MSS.

Editor’s Note

228^b30 It follows: because the difference in heaviness between one piece of earth and another is the difference in the speeds of their natural motion.

Editor’s Note

229^a4 and therefore: the argument is that a change which combines changes of two different species, one after the other, cannot proceed uniformly, but every single change can proceed uniformly (even if it does not always do so).

Editor’s Note

229^a22 in a moment: at ^a27 ff.

Editor’s Note

229^b10 changes, but not variations: at 225^a14–15 Aristotle had suggested that a change from not being pale to being pale could be viewed as a kind of coming to be, i.e. a coming to be of pallor, but presumably this can also be viewed as a variation in the underlying thing which is first not pale and later pale. Here he perhaps has in mind not something which turns pale but something which comes into existence as a pale thing (e.g. a growing mushroom).

Editor’s Note

229^b29 two entities: this apparently means 'two places'. Cf. 225^a3–7.

Editor’s Note

230^a10 no such thing as rest: recall that for Aristotle rest is opposed to variation, and variation does not include coming to be or ceasing to be. Cf. note on 225^a29.

Critical Apparatus

230^b2 Omitting ταχύ‎ (Bostock).

Editor’s Note

230^b3 even when it is not compressed?: Ross's explanation is that corn will grow and ripen even when it is not packed down in the earth (but—e.g.—grown on blotting paper?). This strikes me as improbable, but I have no better explanation to offer.

Editor’s Note

230^b30 it seems to retain what is being left behind: Aristotle is thinking of an extended thing moving from its place, and noting that some part of the thing will still be in the place even after the motion has started (cf. 234^b10–17, 240^b20–31). He goes on to suggest (implausibly) that this part may be said to be at rest.

Share Link

Aristotle

Robin Waterfield and David Bostock (eds), Oxford World's Classics: Aristotle: Physics

Main Text

pg 118V

CHANGE

1. Coincidental changes

The distinction between change and variation

2. The different kinds of variation*

3. Definitions of various terms

4. What counts as a single change

5. How a change is opposite to a change

6. How change and rest are opposites

The opposition between natural and unnatural change and rest

Notes