# Can we really attribute property P to an object with multiple parts which differ with respect to P?

So I've got myself into a tangle with properties. I have quite a strong intuition about something, but a few have told me it's wrong. I'm hoping that either (i) my view is less controversial than they have made it out to be or (ii) it can be modified into something that isn't so controversial. Suppose you notice Craig's eye colour and say:

1. Craig is green

This would be intuitively false, because only a bit of Craig is green: his eyes. So perhaps:

1. Craig's eyes are green

This seems much better, but still, consider that, zooming in, the white of his eyes isn't green, and his pupil isn't either. So:

1. Craig's irises are green

Again better still. But zooming in, Craig's irises display variations of colour; green 19 and green 14 and green N...etc. So we zoom in:

1. The upper left quadrant of Craig's iris is green22 (and so on.)

My point is that we should keep doing this until we cease to discover property variations. Only then can we predicate a determinate shade of green, green22. So we'd say that pigment with structure xyz is green22. If we were to go any further, the property would disappear.

This seems eminently obvious to me. Of course we can talk about Craig's eyes as green, but this is strictly shorthand and unaturalistic, arising because we have a concept of green which we can attribute (my concept of green differs, since I'm colour blind).

tl;dr: I have a strong intuition that it's technically incorrect to speak of an apple as being red since only the apple skin is red and the inside might be white. For an object O with parts p,q&r, how can a (natural/reductivist) property P apply to O unless it applies it applies to all p,q,r? If it applies to only q, why not just say it's a property of q? p&r might not even be the sorts of things to which P can be applied.

Some stuff which I don't know much about but which might help:

• Lewis on naturalistic properties?
• Intrinsic vs extrinsic properties?
• Rives' paper on 'genuine' properties?

Thanks!

• "Technically", we can not attribute property P to an object with multiple parts, even if they do not differ with respect to P. Doing that is called the fallacy of composition. But, as Pasnau points out, "the ubiquitous contemporary philosophical talk of “properties”" has "no more of a fixed meaning" than scholastic form and matter. So there is not much "technical" about it, its use is regulated by loose colloquial conventions. And calling apple red whenever its skin as red is as good a convention as any. – Conifold Apr 27 '17 at 20:13
• I'll try to come back and give an answer, but until then, here's some pointers. Check out Quine's discussion of "Cumulative Properties" in Word and Object. In general, there are several distinctions that are relevant here. For composite objects with parts, you have the distinction between "cumulative" and "distributive" properties. For pluralities, you have "collective" and "distributive" properties. Also, there's a notion that I can't think of a reference for but you might call a "hereditary" property, which would "trickle up". (For instance, "partially composed of water".) – Dennis May 27 '17 at 22:48
• There's going to be a lot of overlap with work in formal semantics, Richard Sharvey's "A More General Theory of Descriptions" is a classic relevant here. – Dennis May 27 '17 at 22:50

I have yet to make up my mind about attributing property P to objects whose parts do not all share the property P or not in the same way. However I disagree that saying Craig is green or Craig's eyes are green is wrong.

Here is my reasoning behind it:

There is a difference between saying Craig's eyes are green to saying Craig's eyes are ALL green.

By saying Craig's eyes are green the only thing you can actually claim is that somewhere in Craig's eyes the color green presents itself.

From the premise "Craig's eyes are green" you cannot make the claims presented below, unless you add additional restrictions to your original premise:

• The size of the green area
• Where in the eyes the green resides
• Whether the area of green is within 1 solid boundary or it might be multiple spots of green separated from each other
• The specific type of green
• ... and so on

Now if you were to have originally said Craig is ALL green or Craig's eyes are ALL green, then you would be absolutely correct that this statement is wrong, unless Craig is one freaky looking dude.

I think there is one further point to be driven out of this discussion.

Saying something is green or red or any color is colloquial. Lets examine this linguistically for a second.

Green and red and colors in general are by definition a property possessed by an object of producing different sensations on the eye as a result of the way the object reflects or emits light.

Simply put colors are properties possessed by objects (what a shocker, right?)

Strictly speaking it is incorrect to say that object X is a property possessed by objects.

So what is a better way of saying this?

Craig contains green or Craig's eyes contain green

It might sound a bit odd at first but its kind of like the difference between "Im good" and "Im well". The first is colloquially used by many many people and accepted as correct; whereas, the latter is strictly by definition correct but most people don't use it in everyday language.

You can also say Craig displays green or Craig's eyes display green. Ultimately using weak verbs like "is" and "are" causes a bit of ambiguity. Strong verbs like "display" or "feel" or "contain" alleviate some of this ambiguity.

Thus as long as you are not restricting the statement by adding quantifiers, saying Craig is green is not wrong. Yes it would be better to say something like "Craig contains green" but due to the colloquial use of the word "is" and its ambiguity as a verb I think it is acceptable.

• This seems like a kind of "reader-response" approach to the question. Can you add sources and integrate this with relevant philosophical literature? – virmaior May 10 '17 at 3:02