Recently I read a comment that most physicalists believe that some threshold of complexity must be surpassed prior to any sentience being exhibited. I've heard of similar ideas a lot, but I've never seen anyone try to defend that view ontologically.

Correct me if I'm wrong, but I believe that the properties that things exhibit are rooted in their ontology. In other word, ontology constrains the range of possible properties that something can possess.

If that's the case, then it seems like there are primarily two possibilities for supporting the complexity theory, depending on whether sentience is a property of things or relations between things:

  1. Sentience results from amplifying some sentient property which things already possess (but is not easily detectable without amplification).
  2. Sentience is a property of complex relations between things. I mean a real property as opposed to being merely an abstract description of a relation (such a crystalline configuration, or the swirliness of a hurricane).

There might be other possibilities I haven't thought of. I thought of a hybrid theory in which it is a property of certain things as well as being a property of a certain relation, but that seemed kind of farfetched. The first possibility which I listed kind of sounds like a Leibnizean idea, so I'm guessing there wouldn't be many adherents to such a view.

I'm especially interested in the second possibility, because it's independent of any informational medium, and it seems to correspond to what people actually mean when they talk about supervenient properties arising from complexity. However, it strikes me as odd that a relation could be said to have any sort of ontology that could give rise to sentient properties (or any non-abstract property, for that matter).

My question: Are there any good arguments to support this theory; or, are there any philosophers who have investigated the ontology of complexity such that it could give rise to sentience?

Edit (in response to comments):

jobermark: Traditionally, there was little or no distinction between properties and powers; properties were seen as being able to bring about effects in people (such as sensations) or other things.

Temperature is a measure of energy, but it's really the energy that brings about physical effects such as transduction in humans. Similarly, acidity is a measure of a compound's potential to bring about certain chemical reactions, so it's a description of its reactivity. For that reason, I'd say that temperature and acidity are descriptive of properties rather than being real properties themselves.

However, whatever definition is chosen, the following should be kept in mind: If the theory explains pain, for example, on the basis of nothing more than abstract relations, measurements or descriptions, it denies that its ontology is anything more than an abstraction. You can suggest such a theory, but the question remains as to whether anyone would accept it. Most people believe that pain is something real that can't be explained away so easily.

Philip Klöcking: It would be more accurate to say that I'm speaking of secondary properties in the Lockean sense, which are those that do not resemble their causes. However, the causes to which they correspond are very real, so I would call those real properties of things. This is as opposed to abstract or conceptual properties which only exist in virtue of the understanding. Sensation are real because they are prior to conceptualization and not the product of it. As Wilfrid Sellars pointed out, "...there is no reason to suppose that having the sensation of a red triangle is a cognitive or epistemic fact." (Although I agree with Sellars on this, it should be pointed out that there's a distinction to be made between triangular form and the concept of triangle: The sensation of triangles is not cognitive, but the conception of them as triangles is.)

I'm speaking of real properties as either phenomena or the correlates of phenomena. This is similar to the way that Kant spoke of the "real in sensation":

"[Phenomena] contain, then, over and above the intuition, the materials for an object (through which is represented something existing in space or time), that is to say, they contain the real of sensation, as a representation merely subjective, which gives us merely the consciousness that the subject is affected, and which we refer to some external object." (Kant, CPR A166/B207)

John Forkosh: Here are five reasons why I'm certain that sensations are not merely conceptual:

  1. True vs. Theoretical — We say things like, "I saw it with my own two eyes," because our sensations provide the ultimate court of appeal for empirical truth. Concepts, on the other hand, don't even have to be perfectly true to be useful. In fact, they are more theoretical in nature. My concept of horse, for example, gets me by until I learn something new about horses, and then I update my concept. For that reason, it's reasonable to assume that almost all of our concepts are to some extent incomplete and inaccurate.
  2. Objective vs. Subjective — Sensations are the means by which we access the objectivity of the world, and so, they are beyond our control. Concepts, on the other hand, are subjective and, in order to be useful, must be subject to the control and revision of the understanding.
  3. Effectual vs. Ineffectual — Sensations could be thought of as being effectual in virtue of their role in engaging our dispositions and inclinations. In addition to that, they are the way that the effects of physical properties are made manifest in humans. Concepts, on the other hand, are tools of reasoning, useful for understanding the effects which come to pass, but they don't really play a role as being effectual themselves, except perhaps in a more abstract way by being intermediate between sensation and judgment. The role of concepts tends to be more descriptive or discursive.
  4. Ends vs. Means — This idea is related to the last in the way that sensations engage our inclinations toward material ends. Concepts play an auxiliary or intermediate role in this process and can be thought of as the means to rationally evaluate the ends which sensation presents us.
  5. Original vs. Derivative — Sensations provide original content, and concepts provide derivative content. Our understanding of empirical concepts is ultimately dependent on sensation to provide their meaning. In fact, even abstract concepts must somehow be rooted in one or more of our fundamental faculties (volition, sensation or moral sense) in order to be meaningful. Concepts can be understood by means of other concepts, but somewhere in the mix there has to be something to ground them in a meaningful way. This is perhaps the biggest problem with claiming sensations are merely conceptual; it leads to an infinite regress. That is to say that representations could only represent other representations without ever having anything which is originally represented.
  • 2
    So you reject the possibility of it being an emergent property of complex systems? I know that this is a slightly frustrating one, as is basically says 'We cannot explain it, but it happens', but afaik it is the most common.
    – Philip Klöcking
    Dec 3, 2016 at 14:12
  • 1
    @PhilipKlöcking. I'm not concerned about how such properties are called, but usually when people speak of emergent properties, they are simply descriptions of relations and have no real ontology. You could, for example, speak of a beautiful face as a property, but that's merely a description of how the whole is perceived.
    – user3017
    Dec 3, 2016 at 14:16
  • 2
    @PédeLeão By this criterion for 'realness' are temperature and acidity 'real properties', or a 'mere abstract description' of the corresponding molecular configurations. (They are the primary examples of emergent properties. And we measured them both before we realized of what they are or are not really properties. So it is hard to claim they are abstracted away from what we eventually discovered to explain them.)
    – user9166
    Dec 3, 2016 at 16:18
  • 1
    @jobermark. I edited the question in response to your comment.
    – user3017
    Dec 3, 2016 at 16:58
  • 1
    So your theory basically asks wether sentience could be understood as a primary property in the understanding of Locke as opposed to being a (mind/description/abstraction) dependent property, i.e. secondary?
    – Philip Klöcking
    Dec 3, 2016 at 18:36

1 Answer 1


David Chalmers takes the reasoning you describe and flips on its head: Levels of complexity can never account for the purely ontological nature of consciousness (his famous "hard-problem of consciousness"), and therefore if physicalism is true, then consciousness must be fundamental, i.e. it has to be a basic property of matter like electrical charge or mass. Otherwise some for of dualism is true. I suppose this what you are alluding to in point (1).

For possibility (2) one can start from James and Russell's neutral monism. From what I understand, Russell arrived at this position from the idea that relations are more fundamental than substance.

The reason why I think Russell's monism is relevant to your question, is that the only way it would make sense to speak of a "real" ontology of properties is if one takes the stance that the fundamental constituents of nature are relations, not substances - see also structural realism. Then it becomes possible for different classes of relations (according to different levels of complexity) to have distinct "real" ontologies. To put another way: if relations are what the world is really made of, then differences between classes of relations are ontologically real differences, not mere abstract differences as you put it.


You must log in to answer this question.