This week at The Junkyard, we're hosting a symposium on Kathleen Stock's recent book:  Only Imagine: Fiction, Interpretation, and Imagination (Oxford University Press, 2017).  See here for Kathleen's introduction.  Additional commentaries and replies will run each day this week.

* * *

Commentary from Patrik Engisch: If Only Imagine...

The focus of my commentary is Stock’s account of the nature of fiction, as developed in Chapter 5 of Only Imagine. To begin with, I take Stock’s account as fitting with a popular line of thought in contemporary philosophy of fiction that claims that in order for one to competently engage with a fiction one should meet the normative demand, unique to fiction, that its content be imagined. I shall call this the Normative View (NV).

There are some issues with respect to how we should understand the normativity at play in NV. A first one concerns its source. Against Walton’s functionalism (Walton 1990) and in agreement with Currie’s Gricean account (Currie 1990), Stock endorses the claim that at least part of the source of NV are authorial reflexive intentions and that anyone engaging competently with a fiction should pick up on them.

A second one concerns the kind of normativity at play in NV. A very demanding interpretation would be to read it in terms of an intentional action done for a reason. For instance, in order to engage competently with Elif Batuman’s The Idiot, one would have to actively imagine its content as a result of having identified that one is provided by the author with a reason to do so. However, this reading is too demanding to be plausible and I shall stick with the following weaker one. The relation between engaging with a fiction and imagining its content is normative in the sense that to engage with a fiction and yet to fail to imagine its content turns out to be sub-competent. Moreover, such a competent engagement need not be something done actively by a consumer. Rather, it plausibly results from sufficient acquaintance with the practice of consuming fiction: once one knows what is fiction and how to interact with it, one is supposed to automatically and rather effortlessly imagine its content.

The central question for NV amounts to the following: Which notion of imagining is relevant for it? Indeed, imagination is a notoriously complex and confused phenomenon and it is far from clear which notion of imagination NV should appeal to in order to make its thesis plausible. Stock is very explicit on that point: she proposes a special notion of imagining, F-imagining, that we should plug in NV to obtain a plausible interpretation. My worry is the following: Stock cannot have both NV and her notion of F-imagining. One of them must go.

To engage with a fiction is a complex process that requires several steps. A very first one is what I shall call the “access step.” Indeed, before one is able of “doing something with a proposition one has in mind” (Walton 1990: 20), one must access its content. Focusing here on literary fiction, this is achieved by deploying basic reading and linguistic skills. For instance, engaging with Batuman’s The Idiot, one finds the opening passage:

I didn’t know what email was until I got to college. I had heard of email, and knew that in some sense I would “have” it. “You’ll be so fancy,” said my mother’s sister, who had married a computer scientist, “sending you e, mails.” She emphasized the “e” and paused before “mail.” (Batuman 2017:3)

To engage with this passage, one must, first, have some command of English and, second, some reading abilities. The conjunction of these delivers access to the set of propositions expressed by this passage.

Second comes what I shall call the “modelling step.” Just like the first one, this second step we are about to describe is not unique to our engagement with fiction. Rather, it is common to our engagement with representations generally and, in particular, to a kind of representations fictions are a sub-kind of: narrative ones. There is something characteristic of representations with a narrative structure in that they, standardly at least, encapsulate information about events that have links together and about recurring participants that have many properties and entertain many relations with each other. For instance, from the above passage of Batuman’s The Idiot, we can extract the following (non-exhaustive) information:

a)      There is a main character, the referent of “I ”;

b)     The referent of “I” went to college at some point in the past;

c)     The referent of “I” has an aunt;

d)     The aunt is married to a computer scientist;

e)     The referent of “I” held a conversation with her aunt about her going to college during which the aunt expressed awe at the fact that her niece was going to college and would use emails;

f)      The aunt is prone to emotional effusion and easily impressed.

These are all contents that are true in the narrative contained in Batuman’s novel and that one can extract from one’s understanding of both the explicit and implicit content of the above passage. Moreover, these contents are entertained, processed, and stored in the consumer’s mind in some specific way: namely, the consumer generates a “mental model” of what is going on in the story (Matravers 2014: 63).

Here are some important claims about the building of such models. First, they require some mental episodes of “taking in” explicit or implicit information extracted from the narrative, such as a)-f). Second, there is a normative relation that obtains between the content of the narrative and the content of the model, in the sense that the model should contain only contents that are part, explicitily or implicitly, of the narrative. Third, for any kind of narrative the following question arises: Which elements of the model are not only true in the narrative but also really true? However, it is quite decisive that this question is separated from the question of which elements should be “taken in” and uploaded in the model of a narrative.     

However, according to NV, if the narrative one is engaging with is fictional, then competent engagement with it requires a third step: imagining its content. According to Stock, this amounts to F-imagining it. My worry is that I cannot really distinguish any substantial difference between the mental episodes responsible for the “taking in” of information extracted from a narrative and the mental episodes of F-imagining postulated by Stock and that, she claims, characterise competent engagement with a fiction qua fiction. Indeed, it seems to me that the mental states occurring as part of the modelling step already fit all of the characteristics of F-imaginings. For instance, reading the above passage of Batuman’s The Idiot, already at the modelling step a reader would be in mental episodes that are propositional, quasi-factual, potentially conjunctive and, finally, that would have contents that would not overlap with one’s beliefs set. Hence, my worry is that F-imagining is not a notion of imagination strong enough to yield a plausible interpretation of NV. In sum, either Stock should stick to her notion of F-imagining and give up NV, or she should stick to NV and make her notion of F-imagining stronger. I would recommend giving up NV.

References

Batuman, Elif (2017). The Idiot. New-York: Penguin. 

Currie, Greg (1990). The Nature of Fiction. Cambridge: CUP.

Matravers, Derek (2014). Fiction and Narrative. Oxford: OUP. 

Stock, Kathleen (2017). Only Imagine. Oxford: OUP.

Walton, Kendall (1990). Mimesis as Make-Believe. Cambridge, MA: HUP. 

Response to Patrik Engisch from Kathleen Stock

In order to respond, necessarily incompletely, to Patrik Engisch’s analysis of my view, I need to emphasise two key points (see Section 1.9 of my book for more detail).

First, I think (roughly, with important details omitted) that the set of fictional truths for a work (its “fictional content”) is determined in relation to what the author of the work intends readers to imagine, and not in relation to what readers actually do imagine. However, in many cases, a reader effortlessly understands what she’s intended to imagine, and so imagines it. In other cases – for instance, so-called cases of “imaginative resistance” (discussed in Chapter 4 of my book) – imagining fails, and a perceptible gap opens up between understanding the meaning of the passage, and imaginatively engaging with it.

There is a close analogy here with the Gricean view of conversation and “speaker meaning”. On the basic version of that view, the “speaker meaning” of a conversation is (again roughly, with important details omitted) what the speaker intends the hearer to believe. It is not equivalent to what the hearer does believe, as a result of hearing the conversation. However, in most cases, a hearer effortlessly understands what she’s intended to believe, and so believes it (via the standard practice of accepting testimony). In other cases – for instance, where the hearer distrusts the speaker – a perceptible gap opens up between understanding the speaker meaning of the conversation, and believing it.

A related point to emphasise is that, on my view, imaginatively engaging with fiction can be very phenomenologically thin indeed. It needn’t involve imagery; it needn’t involve much more than feeling one understands the words on the page, and taking a certain attitude towards them. The presence of this attitude may not itself be “felt”. Again, to make an analogy with the Gricean story about conversation, here too, as a hearer comes to believe what she is told, she may not ‘feel’ that anything is occurring over and above hearing and understanding the words. But nonetheless, there remains a difference in principle between a) understanding the content of a conversation, and b) believing it, even if in most cases the two coincide, and feel indistinguishable from the inside. Equally, there remains a difference in principle between a) understanding the content of a fiction, and b) imaginatively engaging with it, even if in most cases the two coincide, and feel indistinguishable from the inside.

Now, as I understand him, Engisch has at least three worries. One is that I supposedly claim that it is imagining that characterises competent engagement with a fiction, qua fiction. This is what Engisch calls the Normative View, and he ultimately wants to reject it. Actually, I’m not sure that I do think this. It seems to me a probably better characterisation of what I think, that trying to understand what you are intended to imagine characterises competent engagement with a fiction, qua fiction. This is what a competent reader of fiction should do, at a minimum. In many cases, given that authors generally have a good grasp on how to communicate their meaning, and want to make it easy for the reader, there is a short and fairly effortless step from trying to succeeding. However, this need not always be the case: as where certain fictional truths are obliquely implied by an author, in a way which takes close reading of the whole text to detect.

The two remaining worries of Engisch’s effectively concern this distinction I invoke between a) a reader’s understanding what she is intended to imagine; and b) her imagining it. One of these is the thought, also expressed by Matravers (2014), that at stage a), there is in fact “nothing unique to our engagement with fiction” and rather a process “common to our engagement with representations generally”. I don’t think this is true: as I argue repeatedly in the book, though there is of course overlap, there are also several techniques and strategies for interpreting content which are particular to fiction, and which would be inappropriately applied to non-fictional texts.

Engisch’s remaining worry is effectively that, in reading fiction, there’s no real difference between understanding what you’re intended to imagine (what he calls the “modelling” step), and imagining it. But it seems to me that, as noted just now, cases of imaginative resistance bely this. Engisch’s chosen example is one where there is a very smooth transition from understanding to imagining, but not all cases are like this.