Abstract: To understand what non-propositional content is and whether there are any such contents, we first need to know what propositional content is. That issue will be the focus of the first section of this essay. In the second section, with an understanding of propositional content in hand, we will consider representations that fail to have propositional content. In contrast to recent literature, it will be argued that metaphysical considerations concerning what’s represented, rather than linguistic considerations, are a more promising way of establishing non-propositional contents. To keep the discussion containable, focus will be on representational mental states, though many of the considerations can be extended to other forms of representation.
1. Having a Proposition as Content
We don’t typically fear or desire propositions and yet many theorists are attracted to the idea that at least some of our fears and desires are
propositional attitudes. For example, I might fear that the department won’t hire anyone or I might desire that epistemologists apply. These attitudes look to be good candidates for being propositional. Both the desire and the fear are ascribed using a two-place verb flanked by a term for a thinker and a ‘that’-clause. Both are states that are ‘evaluable’ — the fear for realization and the desire for satisfaction. And yet in neither case do we want to say that one must be desiring a proposition or fearing one. So if it is not desiring it or fearing it, what is the relation between a thinker and a proposition when it is true that one, say, fears that the department won’t hire anyone? We need a fuller understanding of what differentiates fearing or desiring a proposition and having a proposition as the content of a fear, desire, or any other mental state. More generally, we need an account that differentiates propositional attitude relations (attitudes that have propositional content) from any other relations we might bear to propositions (such as owning a book that expresses one).
A key feature of propositional attitudes is that they represent things as being some way. When I believe that Berkeley is beautiful, my belief represents a thing — the city of Berkeley — as being beautiful. Propositional desires, fears, and other states are the same in this way. If I fear that the department won’t hire anyone, my fear represents the department as not hiring anyone. And when things are the way they are represented as being, my beliefs are true, my fears are realized, my desires are satisfied, and so on. Because propositional attitudes represent things as being some way, they are representational states with evaluability conditions.
Propositions also represent things as being some way. The proposition that grass is green represents grass as being green. A recent point of contention concerns how it could be that propositions represent and whether we need an account of how it is that they do. For present purposes, let us take it as given (though perhaps in need of future explanation) that propositions are abstract representational entities that are the bearers of truth and falsity.
My belief that grass is green has the proposition that grass is green as its content. Notice that both the belief and the proposition represent the same as each other. And when I fear that the department won’t hire anyone, my fear represents the world as being the very way that the proposition that the department won’t hire anyone represents the world as being. My fear-state and the proposition represent things as being the very same way. I propose that we understand having content in terms of representing-the-same-as.
Let’s work towards a generalization. What relation is it that one stands in to the proposition that the department won’t hire anyone when it is true that one fears that the department won’t hire anyone? It isn’t fearing the proposition. Rather, one must be in a fear-state (rather than, say, a desire-state or belief-state) that represents the same as the proposition. Schematically, one V’s that p iff one is in a V-state that has the proposition that p as its content iff one is in a V-state that represents the same as the proposition that p.
Notice that this suggestion allows us to differentiate fearing the proposition that p from fearing that p. When I (perhaps irrationally) fear a proposition, I need only represent it but I need not represent the same as it. Similarly for other attitudes that look puzzling when we substitute ‘the proposition that p’ for ‘that p’. The reason those substitutions don’t look acceptable is that we shift from talking about attitudes that have propositions as contents (that is, attitudes that represent the same as the propositions that are their contents) to attitudes that simply happen to be about propositions.
This conception of a propositional attitude relation — the represents-the-same-as conception — is compatible with the most popular ways of thinking about propositions and our relations to them. On a classical approach to propositions and the attitudes, propositions are the fundamental units of intentionality and thinkers, sentences, and other representations represent in virtue of standing in appropriate relations to propositions. On such a view, a belief represents as it does in virtue of the believer grasping a proposition. Representation ‘trickles down’ from Platonic Heaven to thinkers. No surprise that propositions and thinks end up representing the same since the mental states represent in virtue of the representational properties of the propositions. But one needn’t adopt such a view. One might prefer to start with the mental states, perhaps by telling a story about teleosematics, asymmetric dependence, or a story about acts of predication. On such views, mental states represent all on their own and those representational properties ‘trickle up’ to propositions that then serve to type, classify, or categorize the mind. But, on a trickle-up view, which propositions classify which states? One very plausible suggestion is that all the states that represent the same as each other are to be classified by making reference to the same proposition and those states that represent differently from one another should be classified by different propositions. For illustration, consider the kind of view recently offered by Scott Soames (2015) according to which propositions are act types. On his view, propositions just are types of representational activities, namely acts of predicating properties of objects. On this view the thinker and the proposition represent the same as each other but the proposition does so in virtue of the thinker. So whether you like a classical trickle-down approach or a trickle-up one, mental states and their contents are brought into representational coordination. It’s surprising how little attention the relation of having content has received. It’s commonplace to see philosophers simply say that we stand in ‘the desire-relation’ to a proposition or that our desires are about or towards propositions, but the first idea tells us very little and the second can’t make sense of having a proposition as content versus being about a proposition. Representing-the-same-as marks a clear improvement and can be adopted by anyone who thinks propositions represent.
By way of brief summary, propositional attitudes have propositional contents and for a mental state to have a propositional content is for it to be a mental state which represents the same as that proposition.
2. Non-propositional States and
their Non-propositional Contents
Under what conditions might we think that there are mental representations that have non-propositional contents? The now forthcoming answer is that a representational mental state has non-propositional content just in case it is a state that represents differently than any proposition. Recall that propositions represent things as being some way. If there are mental representations that represent but do not represent things as being some way, then they fail to have propositional content. If they have content at all (and I think they do), they should have non-propositional content or else they will be predicted to represent the same as some proposition and so be predicted to represent things as being some way. In at least some cases, this would be a bad result.
Suppose you’re playing a guessing game and you’ve been asked to think of a number between one and ten. Mine is seven. When I bring seven to mind, I need not predicate anything of it. Having it in mind, I’m in a very good position to judge whether it is even or odd or to wonder how many other people picked seven, but those are additional things I need not do when I simply bring seven to mind. Such a state seems to be a very good candidate for being an objectual attitude. It is an attitude that is about an object but that does not represent it as being some way. This is further evidenced by the fact that, when I think of the number seven, there seems to be no admissible question of the form: ‘When would my thinking be accurate/true/satisfied/ realized?’
Loving, liking, hating, fearing, and perhaps some cases of wanting also seem to come in objectual varieties. But when I love my new bike or hate the man who stole my old one, I am in a state about something — a bike or a man — but I am not in a state which can be accurate or satisfied. As with thinking of a number, we might ask: ‘Under what conditions would my love be satisfied or accurate?’ This question gets no purchase and that’s because the state in question fails to represent things as being a way.
What are the contents of these objectual states? We saw above that to have a proposition as a content is to represent the same as it. In the present situation, we have states that represent things (but not as being some way) and there are abstract entities that look to have the very same representational profile: concepts of objects. Objectual attitudes have as their contents concepts of the objects they are about. And the relation that one enters into with such a content is just as before: representing-the-same-as. When I think of the number seven, I am in a thinking-state which represents the same as the concept SEVEN. That concept represents the number seven and so does my thought.
States like the ones just discussed — putative objectual attitudes — have received attention as of late. But many of the discussions aimed at motivating their existence have focused on linguistic considerations. On the face of things, when we ascribe attitudes such as loving the bike or thinking of the number seven, we do so using an intensional transitive verb construction. I’m sympathetic to the approaches to intensional transitive verbs that do not ultimately treat them as propositional attitude ascriptions via translation or in deep logical form, but I do not think this is the best way to argue for non-propositional contents. For one thing, language may be an unreliable guide. But for another, with some clever semantic engineering, perhaps an anti-non-propositionalist could convert what look to be non-propositional ascriptions into propositional ascriptions by way of meaning postulates and lose nothing in their theory of language and meaning. A better way to argue is via metaphysical considerations concerning the mental states themselves and, in particular, how they represent. An important role (if not the central role) for contents is to categorize our mental states in terms of how they represent (both ‘trickle-up’ and ‘trickle-down’ theorists can agree here). They capture representational similarities and sameness in a way not that different from the way properties capture similarities and sameness across objects — contents group together various representations that represent the same as each other. In light of this, our theorizing about content should be guided by how various states represent. And it’s hard to see what could be more fundamental to representing propositionally than representing things as being some way. So if there is a good case to be made that a mental state does not represent things as being some way, a theory should not say that it has a proposition as its content.
Are there other non-propositional states besides the objectual? I suspect that there are and we have now seen how to find them: look for mental states (and other representations) that fail to represent things as being some way. Let’s consider two other candidates that have attracted some recent attention.
Concerning what she calls ‘interrogative attitudes’ (IAs), Friedman (2013) says the following:
The IAs look importantly distinct from other familiar wh-attitudes like knowledge-wh: they seem to be attitudes that we can have once we can grasp the question itself regardless of whether or not we can grasp any answers to the question. It seems obvious at this point that questions are the right sort of contents for attitudes like this.
On a now popular theory, in order to have knowledge-wh, one must know a proposition that is an appropriate answer to the embedded wh-question. On such a view, what might look like non-propositional knowing is in fact, upon examination, just more propositional knowing. But Friedman’s observation is that when one wonders-wh, one might not be in a position to grasp any answer to the embedded question. Perhaps one just doesn’t have the needed concepts to answer a question like ‘where can one get a newspaper?’. And yet, one can still wonder where to get one.
At present, I don’t wish to argue one behalf of Friedman. I’m more interested in her observation of a putative fact about representing. To wonder where to get a newspaper is presumably to be in a representational state of some kind. If we follow the knows-wh path, we will predict that to wonder-wh is to be in a propositional state. But Friedman provides reason to think that one who is wondering where to get a newspaper need not be in a position to represent in the relevant propositional way and so there would be an incorrect representational burden predicted if we assigned a propositional content. A question is a better candidate content. And if we adopt my represents-the-same-as suggestion about having content, this is easy to integrate. A question such as ‘who ate the pizza?’ determines conditions under which it is answered. We might then say that to have a question as a content one must be in a state that represents the same as the question. So, for example, when Mary wonders who ate the pizza, she is in a wondering-state that is answered when the question ‘who ate the pizza?’ is answered. She needn’t herself know an answer or even grasp any answers.
One more candidate is worth considering briefly. Recent work on action has been concerned with states answering to ascriptions such as ‘S wants/intends to ϕ’ and whether they are propositional. Clark (2001) and Hornsby (2016) are two proponents of the view that they are not, arguing that ‘to ϕ’ looks to pick out an act or an action and not a proposition. Hornsby suggests that we should have a category of abstracta for acts (since we might intend to perform an act that never gets done). Let’s suppose for a moment that all of this is correct. If it is, then presumably ‘do-ables’ are things that can be done. We might conceive of them as standing for or representing the actions that fall under them. But intentions themselves are also things that can be done, or at least carried out. Following the conception of having content in terms of representing-the-same-as, when one intends, one is in an intending-state that represents the very actions represented by the do-able. If theorists such as Hornsby are right that there are instances of intending where we needn’t represent ourselves (or anyone else), then perhaps a do-able is a better candidate for the content than a proposition that represents some agent and being a way.
To conclude, propositional attitudes have propositional contents and every propositional attitude represents the same as some proposition. A representational mental state has non-propositional content just in case it is a state that represents differently than any proposition. To determine whether a representation has a non-propositional content, we should consider what it represents and in particular whether it represents things as being some way.
Thanks to Lucy Campbell, Craig French, Laura Gow, and Alexander Greenberg for helpful feedback and comments, and to Bence Nanay and his research group in Antwerp for providing an occasion to write this essay. Special thanks are owed to Ray Buchanan with whom I have discussed these ideas at great length.
Ben-Yami, H. (1997) Against characterizing mental states as propositional attitudes, The Philosophical Quarterly, 47, pp. 84–89.
Braun, D. (2011) Implicating questions, Mind and Language, 26 (5), pp. 574–595.
Buchanan, R. & Grzankowski, A. (ms) Propositions on the cheap.
Buchanan, R. & Grzankowski, A. (forthcoming) Propositions as objects of the attitudes, in Tillman, C. (ed.) The Routledge Handbook of Propositions, London: Routledge.
Campbell, L. (ms) Why intentions are not (always) propositional attitudes.
Clark, P. (2001) The action as conclusion, Canadian Journal of Philosophy, 31 (4), pp. 481–505.
Crane, T. (2001) Elements of Mind, Oxford: Oxford University Press.
Forbes, G. (2000) Objectual attitudes, Linguistics and Philosophy, 23 (2), pp. 141–183.
Forbes, G. (forthcoming) Content and theme in attitudes ascriptions, in Grzankowski, A. & Montague, M. (eds.) Non-propositional Intentionality, Oxford: Oxford University Press.
Friedman, J. (2013) Question‐directed attitudes, Philosophical Perspectives, 27 (1), pp. 145–174.
Grzankowski, A. (2015) Not all attitudes are propositional, European Journal of Philosophy, 3, pp. 374–391.
Grzankowski, A. (2016a) Limits of propositionalism, Inquiry: An Interdisciplinary Journal of Philosophy, 59 (7–8), pp. 819–838.
Grzankowski, A. (2016b) Attitudes towards objects, Noûs, 50 (2), pp. 314–328.
Hanks, P. (2015) Propositional Content, Oxford: Oxford University Press.
Hornsby, J. (2016) Intending, knowing how, infinitives, Canadian Journal of Philosophy, 46 (1), pp. 1–17.
King, J. (2002) Designating propositions, Philosophical Review, 111 (3), pp. 341–371.
King, J., Soames, S. & Speaks, J. (2014) New Thinking about Propositions, Oxford: Oxford University Press.
Merricks, T. (2009) Propositional attitudes?, Proceedings of the Aristotelian Society, 109, pp. 207–232.
Montague, M. (2007) Against propositionalism, Noûs, 41 (3), pp. 503–518
Soames, S. (2015) Rethinking Language, Mind, and Meaning, Princeton, NJ: Princeton University Press.
Sinhababu, N. (2015) Advantages of propositionalism, Pacific Philosophical Quarterly, 96 (1), pp. 165–180.
 Editorial note: This paper was the winner of the Annual Essay Prize at the Center for Philosophical Psychology, University of Antwerp, in 2017. The Journal of Consciousness Studies is collaborating with the Prize to publish the winners of this competition in our journal. Please see http://webh01.ua.ac.be/bence.nanay/paw.html for further details.
 Merricks (2009) is a dissenting voice, but the view below allows one to retain that desiring that p and fearing that p are propositional attitudes even in the face of his arguments.
 The literature in this area is growing but see Hanks (2015) and King, Soames and Speaks (2014) for recent discussion and debate.
 For further discussion and development see Buchanan and Grzankowski (ms; forthcoming).
 Such differences might be drawn functionally, in terms of phenomenology, or so on.
 Happily, this view is neutral on the grain (e.g. Fregean versus Russellian individuation conditions) of propositions and mental states. It simply demands that they coordinate on this front when propositions feature as mental contents.
 Why do some attitudes (such as belief) seem to allow the substitution? One option is that whereas states such as fear and desire come in two varieties — the propositional and the objectual — belief only comes in the propositional variety. See King (2002) for an elaboration and defence. Another suggestion is that locutions such as ‘S believes the proposition that p’ are non-literal and should be treated along the lines of ‘S believes Mary’ or ‘S believes the newspaper’. See Forbes (forthcoming) for this approach.
 See Buchanan and Grzankowski (ms) for reasons for thinking that a ‘trickle-down’ view cannot work and a ‘trickle-up’ view isn’t needed when propositions and our relation to them is properly understood. Those issues are beyond the scope of the present essay.
 See also Peter Hanks (2015).
 Objectual attitudes have been argued for by Ben-Yami (1997), Crane (2001), Forbes (2000), Montague (2007), and Grzankowski (2015; 2016a,b).
 For further discussion and debate see Sinhababu (2015) and Grzankowski (2016a).
 Why not simply take the entities these attitudes are about to be their contents? The reason is at least two-fold (and should be familiar from the literature on propositional attitudes). First, objectual attitudes can be about things that don’t exist — I might fear Pegasus. Second, they individuate more finely than what they are about. One might fear Superman but not Clark Kent. See Forbes (2000) for further defence.
 It does not follow on this view that one is thinking of or fearing a concept when one thinks of seven or fears Fido. Rather, one is representing the same as a concept. This mirrors the situation above concerning fearing and desiring propositions when one fears or desires that p.
 See Braun (2011) for additional discussion of questions conceived of as abstracta.
 The terminology of ‘do-ables’ is due to Campbell (ms).