Course Notes – On Essential and Accidental Properties

By Daniel A. Kaufman

Just got finished teaching Copi’s “Essence and Accident” (1) and Cartwright’s “Remarks on Essentialism” (2), in my Knowledge and Reality course, and the issues strike me as important – and interesting – enough, for an edition of Course Notes.

The question of whether the things in the world have essential properties or “essences” is a longstanding one, in philosophy, going back to the Ancient Greeks, who first posed it, in Western thought (as far as we know).  It is still a live issue in academic philosophy today, despite the fact that in the public discourse more generally, “Essentialism,” whether with regard to race, gender, sex, etc., is almost universally rejected, at least in educated quarters.

The question of “essence” is the question of a defining property; one that constitutes the heart of what a thing really is.  Why would one think there are such essences?  Well, from an epistemic standpoint, one might say that some notion of essence is presupposed by the quest for understanding – when we seek to understand something, by way of our many and various forms of inquiry, we are seeking to understand what it really is and not those things that just happen to be true of it.  But there is another vantage point from which to approach the question of essences and that is from the perspective of our understanding of change.

The Ancient Greeks found change a rather perplexing concept.  On the on hand if we say that x changes, we are saying that it no longer is what it was before.  But on the other hand, because we say that it is x that has changed, we are saying that it is what it was before.  So, how do we make sense of this?

For Aristotle, the difference lies first in distinguishing two types of changes: (a) Identity-preserving changes, in which the thing that changes retains its identity through the change, and (b) identity-destroying changes, in which the thing that changes does not retain its identity and becomes some other thing, as a result.

The way to cash out this distinction is to further distinguish between two kinds of properties: (c) Those whose presence is essential to a thing’s identity and (d) those that are not.  And thus, the distinction between essential and accidental properties is born.

A contemporary way of making the distinction is by way of necessity and specifically, de re necessity, rather than de dicto (‘de re’ meaning “about the thing” and ‘de dicto’ meaning “about what is said”).  To illustrate, consider the following statement:

(e) Necessarily, x is F.

Read de dicto, what this says is:

(f) The statement ‘x is F’ states a necessary truth.

Read de re, what this says is:

(g) x has the property of being necessarily-F.

The idea among some contemporary philosophers has been that we can identify the essential properties of things by seeing which de re necessities are true.

There is, however, a Problem, and it is one that applies, no matter how you arrive at your list of essential properties, namely, descriptive relativity.

One might think that it is an essential property of markers that they write and only an accidental property that they are brown.  Thus, if I re-ink my brown marker, so that it has black ink rather than brown, I will only have changed an accidental property, and the change, consequently, is identity-preserving.  But notice that this is only because our interest in markers is commonly in their writing and not in the colors in which they write.  This is reflected in the fact that we have a word for markers but not a word for brown markers or black ones.  But suppose that we were interested in the colors of markers, such that we had distinctive words for markers of different colors.  ‘Brarker’ for brown markers, ‘Blarker’ for black markers, ‘Grarker’ for green markers, etc.  Imagine even that we had no interest in markers, generally, and that consequently, there is no generic word for markers.  In that case, being brown would be an essential property of a brarker and thus, swapping its brown ink for black would constitute an identity-destroying change.  A brarker would have ceased to exist and a blarker would have come into existence.

We can raise the same problem for essences, modally construed.  As Quine observed, in Word and Object:

Mathematicians may conceivably be said to be necessarily rational and not necessarily two-legged; and cyclists necessarily two-legged and not necessarily rational. But what of an individual who counts among his eccentricities both mathematics and cycling? Is this concrete individual necessarily rational and contingently two-legged or vice versa? Just insofar as we are talking referentially of the object, with no special bias toward a background grouping of mathematicians as against cyclists or vice versa, there is no semblance of sense in rating some of his attributes as necessary and others as contingent.

The example is not ideal, but don’t get lost in the details of whether one might be a paraplegic cyclist or a hot-headed mathematician.  If our interest in Oscar is as a cyclist, then being two-legged is an essential property and being rational, an accidental one.  But if our interest in Oscar is as a mathematician, then being rational is an essential property and being two-legged an accidental one.  If we ask what Oscar’s essential properties are, independently of these and any such descriptions, the question would seem impossible to answer.

You may be wondering what the problem is.  Why does it matter if properties are essential or accidental, only under a particular description? From the standpoint of understanding change, it may not matter much.  Just as what counts as an essential or accidental property is relative to a description, what kind of change has occurred is relative in precisely the same way.  There is no absolute or description-free conception of change and no absolute fact as to whether a change has been identity-preserving or identity-destroying.

It is, however, a serious problem from the epistemic standpoint.  For those who think that the ultimate point of inquiry is to understand the true nature of things – what they really are, not what they are relative to various interests or under various descriptions – the relativity that I’ve described and which seems inevitable, spells trouble.  Indeed, I would maintain that the absence of absolute essences – the fact that it makes no sense to speak of a thing’s true nature, independently of all interests and descriptions – is connected to more general, far-reaching questions, like those pertaining to the Realism/Anti-Realism dispute.  After all, if we only speak of a thing’s essence, relative to various interests and descriptions, how likely will it be that we can speak of things at all, other than in terms of such interests and descriptions?

Fortunately, Realism and Anti-Realism are the subjects we are going to take up next, in the course, by way of Quine’s and Putnam’s and Davidson’s work on conceptual and ontological relativity, so if this issue is keeping you up at night, you won’t have long to wait.


(1)  Irving Copi, “Essence and Accident,” The Journal of Philosophy, Vol. 51, No. 23 (1954), pp. 706-719.

(2) Richard Cartwright, “Some Remarks on Essentialism,” The Journal of Philosophy, Vol. 65, No. 20 (1968), pp. 615-626.

(3)  W.V.O. Quine, Word and Object, The MIT Press, 1960, Ch. 6.







21 responses to “Course Notes – On Essential and Accidental Properties”

  1. davidlduffy

    Interesting. I suppose, as you imply, that part of the problem here is the timelessness of the definitions. I read that “box” in temporal logics translates to “henceforth” or “hitherto”. Oscar was a mathematician-cyclist until he became demented and had a leg amputated – henceforth he lost an essential part of his identity.

  2. Thomas Jones

    Just a note of thanks for providing another of these “Course Notes.” Looking forward to the next entry on Realism and Anti-Realism.

  3. TJ: You’re most welcome. I’m glad that some of our readers find them useful. I’ve always thought of this sort of online work as being part of my work as an educator, so I’m glad to spread what (little) I know around.

  4. Hi Dan, like Thomas I appreciate this even if I might have little to add. In this case I have to mull over how much I require the concept of essence to employ realism.

    For example, to believe that there is an object by which I write or draw using a liquid which dries on a surface… does that really hinge on essentialism?

  5. brodix

    Without change, nothing exists. With change, nothing exists forever. Change is essential.

  6. Thanks Dan,

    It seems like so many of these issues in philosophy are related to the need or the desire to seek absolute foundations. Is this not a similar issue to your recent topic regarding the focus on rationality as well as on the True, and the Good. I have been reading a lot of John Dewey lately, working my way through ‘Experience and Nature’, some of his other original essays, and some papers on his work. It seems like much of his thought was a reaction against these foundational trends. Of the western philosophers I have read his writings resonate most strongly for me.

    I know you don’t mention him much Dan, although on a number of issues it seems to me you both hold similar oppositions to the primary consensus in analytical thought. Based on your interactions with Massimo I am guessing that Dewey’s commitment to continuity of inquiry might be one reason you hold his views in disfavor. Is that right, and are there other reasons?

    Thanks, I value all your responses and would like some checks against my current Dewey fandom 🙂

  7. Dan,
    Very interesting.

    I’m reminded somewhat of the “White Horse Paradox,” expounded by Kung-sun Lung, of the School of Names, in China around 300 BCE. The preservation of the text has been inadequate so certain terms are unclear. The American reading of it, as one can find at SEP, tends to find it a bit of sophism, which hardly does it justice; the Eastern reading tends to project Taosim into it, which is hardly better. *

    The fundamental issue in the readings is determining whether, In suggesting that “’a white horse is not a horse’ is an acceptable proposition,” Kung-sun Lung is ambiguating kinds and identity, pulling a trick of semantics for purposes of edification, or illustrating ‘use/mention’ differences between talking about things, and talking about talking about things.

    Here’s how I read it: if a property is essential to a sub-class of a given larger class (wherein the property is deemed accidental), does that so mark the sub-class as separate from the larger class? Logic says it shouldn’t; but in making determinations about the sub-class, it certainly seems that we’re headed in that direction.

    “Master Kung-sun replied, ln calling for a horse, either a yellow or a black horse will do, but in calling for a white horse, a yellow or a black horse will not do.’”

    As we can see, there’s pragmatic value at play here; ‘I don’t want just any horse, I want a white horse.’ Which gets thornier still, when we start to individuate – imagine in France, 1810: “I don’t want just any horse, I want the white horse Napoleon rode!’ Now, if Napoleon is the one demanding this, we can see why his underlings may want to pay attention to the ‘class of one white horse ridden by Napoleon.’

    This also raises problems concerning how we view individuals. Imagine Bob, a cyclist mathematician, who has had a catastrophic accident involving his right leg. The doctors say the leg must come off; Bob insists, ‘I can’t imagine myself without a right leg!’ We want to reassure him that loss of the leg does not change who he is. But it certainly changes what can be said of him following, e.g. (without a prosthesis and retraining) we can no longer say he is a cyclist. And that will affect how he sees himself.

    But I was also reminded, while reviewing the White Horse Dialogue: We say of the class ‘horse’ that color is an accidental quality; but that’s not precisely true – any actual horse must have some color. So while any particular color may be accidental; color as such is essential to it. So too any visible entity; so too Bob. There are a number of ‘essential qualities’ needed for any entity to exist in our universe of 3+1 dimensions, and still others such that we can sense them – most of which we don’t even think twice about. Finding ways to focus on those we need to articulate and negotiate would appear to be an essential property of inquiry.

    * I prefer the discussion by Thompson at , which includes translation of the text, from whence I drew my quotes.

  8. Thanks EJ,

    I believe Zhuangzi references the ‘white horse’ in the 2nd inner chapter. I’m going to check out your link.

  9. … period of mulling ended…

    Hi Dan, so I gave this matter some thought and am dubious essentialism impacts metaphysical/ontological realism.

    To discuss the essence of things is to discuss how we speak about things. It seems concerned with definitions, categorizations, and their usages: how they can be applied to “things”, or how “things” can fit into the definitions/categories we make.

    To argue the utilitarian and/or limited nature of our definitions impacts the veracity of metaphysical/ontological beliefs, or undercuts the possibility things have certain properties (whether they are epistemically accessible or not) appears to conflate communication or representation of things with the nature of things. While essences arguably exist for definitions, this thing —> (indicates pointing to a thing) is not bound by any particular “essence”.

    For example, whether this object —> is a “pencil” may very well be language/intention bound. But whether this —> will be able to produce lines on the surface of another thing here —>, and how this feat is believed to be carried out (adding pressure while pressing one to the other with parts of the first smearing across the second), is not.

    We may try to codify or represent our experiences within language, which by its nature may be limited and vague (or at least not all-encompassing), but the question of if our knowledge about the world or the nature of the world itself is undercut in any meaningful way (in all cases) by the limits of our tools of communication ignores that there are other ways than language of interacting with and holding/expressing beliefs/knowledge about the world.

    That we can demonstrate drawing without being able to pin down the exact true essence of “pencil” or “paper” or “graphite”, suggests that we can at least talk usefully about about this —> being a pencil that we “know” will allow us to draw —> on paper —>. And if we have been able to build a pretty good microscope (itself requiring several assumptions about the world to be relatively accurate), that these lines —> coming from the pencil will be seen to consist of these objects —> which we can if we choose if we want (for convenience) to call sheets of graphite.

    That both the pencil and paper in another context may also turn out to be great “kindling” does not put anything about their nature in question, even if multiple definitions (and so multiple sets of essences) end up applying equally well to them.

    Our ability to successfully construct new objects and abilities (for example a microscope for which there was no prior definition, so we create an essence?) based on an increasingly sophisticated sets of beliefs about the world generates an implicit (if non-lingual) argument for the relative validity of those beliefs (regarding the nature of the objects we experience).

  10. db. The thought was something like this: Essentialists do not believe that essential properties are merely psychological or conventional — matters of how we decide to classify objects — but represent the heart — the *reality* — of what something is.

    Essentialism strikes me as a view that one can only hold as a metaphysical Realist. I don’t see how someone like, say, Nelson Goodman, could be an Essentialist.

    Now, a Realist, when confronted with conceptual relativity arguments may sometimes retreat into what I call a “reality as generic ‘stuff’ ” view. A kind of crude version of Kantian Noumenalism. The trouble is that there are good reasons to think this view is incoherent — some of the best are found in Davidson’s paper, “On the Very Idea of a Conceptual Scheme” (which I will be teaching in my course, soon).

    The realist, then, cannot say that independent reality just consists of generic stuff. Independent reality consists of things, with properties. But that means that independent reality is already broken up into categories. And given that it is supposed to be *independent* reality — we are talking about Realism, after all — that would seem to me to require real Essences.

    It seems to me, then, that Realism requires some notion of real Essences. And that a conventionalism about Essences is natural to an anti-Realist perspective.

  11. Hi Dan, thanks that was pretty helpful and potentially challenging.

    I definitely got the idea this is what you were getting at (and I agree):

    Essentialists do not believe that essential properties are merely psychological or conventional — matters of how we decide to classify objects — but represent the heart — the *reality* — of what something is.

    I also agree with your claim that essentialism can only be held by a metaphysical realist. Where we might start disagreeing is that I would caveat that to “a certain kind of” metaphysical realist. And the kind of realism I advocate would not accept the concept stated above.

    But before I can go further I’d be interested in knowing how you might differentiate essentialism (or if you do) from Platonism. For me they seem very closely related (the above quote seems to describe a form of Platonic realism), and problematic for about the same reasons. To be honest, it is why I never spent much time worrying about essentialism. But maybe I was being too hasty/shortsighted.

    Without going further I suspect you would view my position as the retreat you described, though I reject (as I did in an earlier thread) Kantian Noumenalism. That is incoherent, but I think an inaccurate model for what I want (or need) to claim. It will likely revolve around differences on what we mean by or how we conceptualize “generic stuff”.

    I would reject that having a specific set of properties means that things in the world are “broken up into categories”. That appears to conflate the nature of properties with language we (must) use to convey meaning to each other about those properties… which smacks of Platonism.

    If you are going to be giving the Davidson paper a go here, we can wait until then to see how I might deal with the challenges it presents. I think (and hope) I am coming at this from a slightly different position than you have anticipated.

    Of course, it will be interesting to see if you can shift me to anti-realism based on this line of argument.

  12. db:

    First, I am *not* trying to convince you to be an anti-realist.

    Second, I don’t see why Essentialism need entail Platonism. The Essential property of a thing can be a natural property. Even a property like “being used for writing” is not particularly Platonic or at least, need not be construed that way.

    Third, if independently existing things have independently existing properties, then that means that there are independently existing kinds. How could there not be? Say that there are five independently existing things that independently possess the property, F. Then there independently exists a kind — the F-kind.

    I will be teaching Quine’s “Ontological Relativity” and Davidson’s “On the Very Idea of a Conceptual Scheme” as well as excerpts from Goodman’s “Ways of Worldmaking” in the next two weeks (our last two weeks), and will probably post two more Course notes covering them all, in some way.

  13. Seth:

    Sorry it’s taken me so long to reply. The reason I don’t talk much about Dewey is that I know next to nothing about his work. Analytic Philosophy programs to a great extent — or at least when I was in graduate school — largely ignore him, as they do Peirce. James is about the only American Pragmatist you are likely to get, if lucky.

    I consider it a substantial hole in my education and suspect it narrows my imagination. I am therefore always happy to have someone bring him into a conversation, in order to expand it.

  14. Hi Dan, I wasn’t accusing you of trying to push anti-realism.

    Your arguments happen to attach essentialism (something I disagree with) to realism (something I have held for a long time).

    Whether it is your intention or not, if you successfully make the case that essences being real is “essential” to realism, or at least that realists have no effective way out of that attachment (beyond some unsatisfying noumenalist account), then you have delivered a case against holding realism (for me). It would force me into a review of a position I haven’t felt the need to recheck in a long time.

    That would be interesting! 🙂

  15. Hi Dan, I kept trying to write a response to your points but it keeps spinning out of control. Way too long.

    So let me ask some short questions. If independent natural properties do not in fact exist, why is there change? How is change possible? If there is only one class of thing then there seems no reason for that thing to change, a mechanism by which to manifest change, or an ability to see change.

    The existence of change and interaction seems to demand the existence of different natural properties, regardless/independent of our interests. And so it is not clear to me how anti-realism avoids accepting the existence of independent natural properties, even if they may be less in number than what realist accounts might demand.

    This is wholly separate from how anti-realists/realists treat the generation of definitions/models by which we impose categories and kinds on the world in order to understand it. It seems to me that anti-realism denies that definitions/models can be accurate to the world, and your argument here uses the inherent limitations of “essences” to deny the possibility of accuracy.

    While I would argue that realism allows for a variable degree of accuracy between definitions/models and the world. The limitations of essences may restrict our ability to describe a thing in toto, but cannot deny our ability to accurately describe aspects of a thing (in some cases in toto). Nonverbal manifestations of our beliefs in the world (where they are successful) raise challenges to anti-realist denials of accuracy based on limits of descriptions, and is devoid of relying on “essences”.

    By the way, your course sounds pretty nice!

  16. A number of comments here raise interesting issues but I was particularly interested in the exchange between Daniel Kaufman and dbholmes.

    Dan wrote:

    … The realist, then, cannot say that independent reality just consists of generic stuff. Independent reality consists of things, with properties. But that means that independent reality is already broken up into categories. And given that it is supposed to be *independent* reality — we are talking about Realism, after all — that would seem to me to require real Essences.

    It seems to me, then, that Realism requires some notion of real Essences. And that a conventionalism about Essences is natural to an anti-Realist perspective.

    The last sentence I agree with. But I also think that there is no reason to believe that a ‘scientific realist’ (as I and I think dbholmes would understand the term) needs to worry that they might by virtue of their ‘scientific realism’ be unwittingly committing themselves to some kind of essentialism.

    Dan’s claim that the (scientific?) realist believes in an “independent reality” which consists of “things with properties” is, it seems to me, a very misleading (and metaphysically loaded) way of putting the (scientific) realist point of view, at least as I understand it. (I’m uncomfortable with the terms ‘realism’ and ‘scientific realism’: I see the latter as shorthand for a normal scientifically-based view of reality.)

    I guess this discussion will be carried forward into a new post. If it is I will be interested to see what is said about these matters; but my guess is that most reflective, scientifically-minded people know well enough the basic (minimal) assumptions and principles to which they are committed and will not be too concerned about how their position might be labelled (so long as it is understood for what it is).

  17. Mark, this discussion is about Metaphysical Realism, not Scientific Realism.

  18. Dan

    I am aware of the distinction. I know something about that late-20th century debate within analytic philosophy. But (perhaps mistakenly) I was assuming that dbholmes was arguing more for something like a commonsense or scientific realism.

  19. The arguments posed are only arguments against metaphysical Realism. They are not arguments against either common sense or scientific realism.

  20. Hi Mark and Dan,

    …there is no reason to believe that a ‘scientific realist’ (as I and I think dbholmes would understand the term) needs to worry that they might by virtue of their ‘scientific realism’ be unwittingly committing themselves to some kind of essentialism… I was assuming that dbholmes was arguing more for something like a commonsense or scientific realism.

    Mark, you have a correct read of what I was trying to get at. That’s why I said (to Dan)…

    I also agree with your claim that essentialism can only be held by a metaphysical realist. Where we might start disagreeing is that I would caveat that to “a certain kind of” metaphysical realist. And the kind of realism I advocate would not accept the concept stated above.

    … which tried to distinguish between forms of realism. This is why I took your (Dan’s) response to that comment as maintaining essentialism as true for all forms of realism, with no exclusion for scientific realism (as well as other “softer” forms).

    That said, my understanding of scientific realism (though I am in no way an expert) is that there are some metaphysical commitments such that it can make sense to discuss an “independent reality” (specifically mind-independent) and “things with properties”. So I was not completely thrown off by Dan’s use of these terms. My own replies have used them. But yes, we have to be careful with such terms, including the level of commitment held (and where it applies).

    Perhaps most of this has been cleared up, if all claims made were for “hard” metaphysical realism, and excludes other forms. Basically that agrees with my quote above.

    I am still interested in understanding how anti-realists (or anyone) can escape belief in the existence of differing properties. Whether they are ever accessible to us or not, including the possibility we can never get to them due to limits of our language use, it is clear there are differences in experiences which necessitate different “properties” in the world. If there is only one property (or no property) then whence comes experience?

  21. Hi Dan,

    Thanks for the response. I think you would enjoy delving into Dewey, although I’m a newbie and certainly not a good candidate to make his case. He is not the easiest to writer to grasp in short readings, I think it takes cumulative effect when he applies his thought across multiple domains. I’m guessing you would be sympathetic to much in his approach, but would also have problems with some aspects. In all I think you would find the time worth your while. ‘Nature and Experience’ might be a decent place to get a read on his mature thought.