Shen-yi Liao is Associate Professor of Philosophy at University of Puget Sound. He is interested in the imagination, but also in too many other things.
A post by Shen-yi Liao
Overview
Is imagining continuous with remembering? In recent years, this (dis)continuity debate has received much attention from philosophers of memory (including on The Junkyard: here and here. In a collection of forthcoming works—
“What Sort of Imagining Might Remembering Be?” (“Constructive”),
“Remembering, Imagining, and Memory Traces: Toward a Continuist Causal Theory” (“Memory Traces”),
“Propping up the Causal Theory” (“Causalist”), and
“Remembering and Imagining: The Attitudinal Continuity” (“Attitudinal”)
—Peter Langland-Hassan shows how philosophers of imagination can contribute too. With his characteristic analytic acumen, Langland-Hassan’s forthcoming works clarify and advance the debate about the relationship between imagining and remembering.
In this debate, nearly all agree that ‘remembering’ refers to episodic memory.[1] But what does ‘imagining’ refer to? Langland-Hassan rightfully points out that philosophers of memory are not always precise about what they mean by ‘imagining’. And this is a serious problem because the nature of imagination is highly contested. On the rough characterization that to imagine is to represent without aiming at things as they actually, presently, and subjectively are, there are at least three dimensions of variation: modal, temporal, and subjectual. Most relevant to this debate are imaginings along the first two dimensions: specifically, episodic counterfactual thoughts and episodic future thoughts.
As Langland-Hassan characterizes the (dis)continuity debate, the continuist claims that episodic memory and episodic counterfactual and future thoughts are all instances of imagining, and the discontinuist denies this claim.[2] Borrowing a distinction from Neil Van Leeuwen (2013), which has also been employed by Sarah Robins (2020), Langland-Hassan disambiguates ‘imagining’ into three possibilities: imagistic, attitudinal, and constructive. He then argues that the most plausible interpretation of the (dis)continuity debate concerns whether episodic memory and episodic counterfactual and future thoughts arise from the same process of constructive imagining (“Constructive”; see also Wiltsher in press).
Although it is standard to align continuism with simulationism and discontinuism with causalism, Langland-Hassan argues for a continuist causal theory. This theory is continuist because episodic memory and episodic counterfactual and future thoughts all arise from constructive imaginings with memory traces, but the memory traces themselves are more like props for reenactments, and not like replays of recordings (“Memory Traces”). However, this theory is also causalist because it demands genuine episodic memory to have an appropriate causal connection to the original episode. Importantly, this appropriate causal connection need not be via memory traces themselves, but can be via other metacognitive mental states involved in a remembering episode, such as a related belief or a fluency feeling (“Causalist”).
In addition, Langland-Hassan also considers another interpretation of the (dis)continuity debate that concerns the attitude of imagining, which is its most common sense in philosophy of imagination. In this debate, Robins suggests that seeming to remember—the non-factive attitude that includes but also goes beyond remembering—and imagining are distinct because of metacognitive mental states, such as “what the person engaged in these mental states takes themselves to be doing, what they’d report about their mental state if asked, and what they would go on to think or do as a consequence of this mental state”. In contrast, Langland-Hassan suggests that memory and counterfactual and future thoughts are all mental states of the same attitude: imagistic judgments (“Attitudinal”).
This post mostly exists as a letter of recommendation: go read these! I am sympathetic to Langland-Hassan’s overall picture, and I think it has helped me to better understand the relationship between imagining and remembering—and, even more importantly, the many moving parts in this relationship. It has given me some ideas for understanding the (dis)continuity debate, and how trying to understand this debate can be useful for theorizing about imagination.
Debate Space
Langland-Hassan approaches the debate from an abstract starting point (“Constructive”). In general, he says, to call events A and B continuous is to affirm that they belong to the same kind C. From there, it is a matter of filling in the variables, such that A = episodic remembering, B = episodic counterfactual and future thinking, and C = imagining. There seems to be clarity to this abstraction; we can Venn diagram our mental lives neatly into circles.
But there also seems to be an industrial-strength realist presupposition behind this appearance of clarity: that our mental lives really do neatly break down to events type A, B, C that can be so Venn diagrammed. To be fair, this industrial-strength realist presupposition is not Langland-Hassan’s own invention; much of the (dis)continuity debate really does seem to proceed in these terms. Elsewhere, I too have exhibited this boxological impulse.
While I still see value in these diagrammatic abstractions, I don’t think they should be interpreted in industrial-strength realist terms. Instead, following Daniel C. Dennett (1991), I think of them in mild realist terms as patterns: our mental episodes are stars, and there are constellations to be found, but these real yet noisy patterns are necessarily viewed from specific perspectives that depend on its capacities and interests. ‘Episodic remembering’ is an efficient description of a cognitive phenomenon, ‘episodic counterfactual and future thinking’ is an efficient description of another cognitive phenomenon, and—the continuist affirms—‘imagining’ is yet another efficient description of a cognitive phenomenon that more or less encompasses the first two.
When it comes to folk psychological concepts, there always seems to be at once a theoretical aspect and a metacognitive aspect. We are trying to construct theoretical hypotheses, but we are also trying to understand ours and others’ minds. And inevitably there are different patterns that we see, from these two different perspectives that come with different interests.
César Schirmer dos Santos, Christopher Jude McCarroll, and André Sant’Anna (in press) suggest that continuists and discontinuists are engaging in a partly semantic debate about the meaning of ‘remembering’. In particular, they say that causalists qua discontinuists want to define ‘remembering’ as requiring an appropriate causal connection, and simulationists qua continuists do not. In addition to rejecting the alignment of continuism with simulationism and discontinuism with causalism, Langland-Hassan also rejects the framing of the debate as being partly semantic: “Why isn’t this then a straightforward dispute about the nature of remembering? What makes it meta-linguistic?” (“Memory Traces”).
I am not sure. I think the less realist we are about psychological concepts, the more plausible it is to see debates like these as being partly semantic. The rejection of industrial-strength realism implies that there is no (perspective-independent) the nature of remembering; there are different patterns that we can pick out, and which ones we in fact pick out depend on our capacities and interests. Insofar as we engage in metalinguistic negotiation, it can be about which interests are most salient in a given context: for example, the interests of constructing theoretical hypotheses or the interests of understanding ours and others’ minds.
Memory or Metamemory
That said, I want to suggest that insofar as there is metalinguistic negotiation going on in the (dis)continuist debate, it is not centered on the inclusion of a causal requirement. Instead, I think that the negotiation is about the boundaries between memory and metamemory—the metacognitive aspects of a remembering episode such as a related belief, a fluency feeling, and so on.
As an example of the work that this boundary does, consider the influential view developed by Kourken Michaelian (2016). Even as a paradigmatic continuist, he accepts that there can be metacognitive differences between remembering and imagining. Indeed, it is part of his overall account that remembering co-evolved with metacognitive aspects of remembering, such that there is “‘preestablished harmony’ between the way in which memory tends to simulate episodes and the shape of the episodes that we actually experience”. The boundary between memory and metamemory is, in other words, how Michaelian can have his cake and eat it too.
Langland-Hassan’s continuist causal theory exploits this boundary too, in a different way. The theory is continuist when it comes to remembering, but causalist when it comes to meta-remembering: the appropriate causal connection need not be satisfied via remembering, narrowly understood, but via the metacognitive mental states, such as a related belief or a fluency feeling.
But it feels like the same discontinuists qua causalists who find Michaelian’s continuism dissatisfying would also find Langland-Hassan’s version of causalism dissatisfying. It seems important for them that the appropriate causal connection falls on the side of memory, and not metamemory. And, for them, that is the source of the discontinuity between remembering and imagining.
Or, these discontinuists might accept Langland-Hassan’s version of causalism, but deny that it is continuist on an interpretation of the (dis)continuism debate such that it is not about remembering and imagining, narrowly understood, but also meta-remembering and meta-imagining too. Indeed, that seems to be the driving thought behind Robins’s suggestion that seeming to remember and imagining are distinct because of differences in their respective metacognitive mental states.
Back to Imagination
As the reader of The Junkyard knows, the nature of imagination is highly contested. I can’t help but wonder whether some of the ideas from the (dis)continuity debate can also be applied to debates about imaginings. Sometimes, in search of clarity in contradistinction, the boxology debates seem to carry an industrial-strength realist presupposition, even when it might not be reflectively endorsed. And sometimes, it feels like we are engaging in some metalinguistic negotiation about the meaning of ‘imagining’, depending on the interests of our perspectives.
Notes
[1] There is actually a further complication about the difference between remembering and memory, partly corresponding to a distinction between a cognitive process and a cognitive state. I am ignoring this complication for now because this ambiguity, unfortunately, seems also inherent to the ambiguity of the (dis)continuity debate itself. See also note 2.
[2] Langland-Hassan explicitly aims to interpret the (dis)continuism debate in an idealized way such that the two sides are not merely talking past one another (“Constructive”). By contrast, Robins (2020) interprets the two sides to be in fact talking past one another, to an extent: “For discontinuists [...] the focus is on episodic remembering as an occurrent mental state. [...] The continuist, in contrast, is focused on the episodic memory system—the neurocognitive structure that gives rise to episodic memory”. In addition, Schirmer dos Santos, McCarroll, and Sant’Anna (in press) interpret the two sides as engaging in a metalinguistic negotiation over the concept of remembering. Finally, David Colaço (in press) suggests that theorists who define the concept of memory differently can be interpreted as offering different hypotheses about the cognitive phenomenon of memory.
Thanks to Peter Langland-Hassan and David Colaço for discussion.