👀
20.01.2026 00:53
👍 0
🔁 0
💬 0
📌 0
👀
The logic holds for any imagined object/scene failing to overwrite basic vis features in a retinal image but the extremity of the comparison often gets people to go "Oh, I understand the "seeing" he means now". Very happy to explain the QS account over Zoom too!
I think MI totally affects EVC, but definitely doesn't reactivate like a retinal im. so there's no contradiction: if MI modulates EVC, but not reactivates, you can decode from EVC in fMRI, but won't see cross-decoding in MEG/EEG to early representations, only mid/later ones (which is what we see).
Exactly! I want to do the same: "Here's a list of possible experiences... choose". Would be keen to collab. Still some introspective challenges (e.g. people saying MIs are just like retinal images until "So visualising a pitch black room is visually identical to walking into a real one?") But yes.
I think detecting atypicality is totally valid as long as we actually define EXACTLY what we mean by "seeing" (I definitely don't reject the possibility of aphantasia), but no one's done this aptly. The QS account tries, though. It just feels like we've drawn the cart before the horse otherwise
Definitely tricky but I'd predict sound-guided imagery and dreams would be preserved (but the latter would be experienced in a very different way, closer to hazy mental imagery than a retinal image experience)
No guarantee the QS account is actually typical (that's an empirical question), but it's one possible solution to imagery being different to dreams and retinal images while still feeling "seen". To my knowledge, there's no other explanation that holds up to these critiques: philarchive.org/rec/AMYIV
Fair - could be uniformly, not normally, distributed, but the quasi-sens account covers that variation. I regularly meet people who say they see their imagery as clear as day, and others who say they see nothing, but after enough talk, both often agree that their experience matches a QS experience.
True, lesions can be fuzzy, but there are multiple case studies and some seem to involve complete destruction and complete blindness - they cannot see anything in front of them, but they say their mental imagery is the same. TMS could touch on the same question, but its effects are much smaller
Regarding V1, to me it seems more like the least, rather than most, relevant part of the vis hierarchy for imagery (see pic and also doi.org/10.1016/j.ne...). This fits the quasi-sensory view, but I'm honestly not sure if I'm communicating it well enough given the backlash I get on it.
I then avoid giving them much weight because of (1) criterion/language confounds and (2) if someone doesn't think they can do imagery, even if they actually can, they'll probably do nothing much during a study's imagery period (could that explain most positive aphantasia findings?)
Using aphant studies to determine what typical imagery is before correctly defining typicality seems circular to me: it presumes that aphants have accurately ascertained ground truth of what typicality is (because they are rejecting it) despite this being the very thing we are trying to figure out.
My goal is to determine what "seeing" typical imagery truly means. I assume the lesioned folk had typical imagery before their injury (there's no reason to suspect otherwise), so if they say it's unchanged/just as vivid, that lets us make conclusions about what typical imagery probably involves.
Am curious about this debate because I feel like lesion studies would be fairly conclusive evidence that V1 is not necessary for mental imagery anyway. It can still be depictive without EVC (but "depictive" != "identical to a retinal image experience"). From doi.org/10.31234/osf...
No doubt that people's imaginations have diff creativity, content, emotion, etc, but this is why the visualising a black field is good: it controls for all these things, because none of them seem to explain variance in the latent factor. E.g. see the Kind 2017 link
Maybe you're taking vividness diffs to mean any perceptual diffs, & I'm taking them specifically to mean a diff in the latent variable separating retinal and mental image perception. This is why the variance in your survey can be real without it being experienced differently on that latent variable.
(PS the talk summarises what people who think they have typical imagery think they do. By reference experience, I meant “Is your mental image of X EXACTLY identical to experience Y?” e.g. is visualising a black field EXACTLY identical to the vis experience of a real one?)
If this seems like I’m talking complete nonsense, please try osf.io/preprints/ps... (the NoC paper's definition was way too vague). I seriously think this is just a misunderstanding of quasi-sensory vs sensory experience. Always happy to take this to a one-on-one discussion too.
Watched the whole 1.5 hrs. You’re gonna hate me Reshanne, but virtually every person’s description here, from aphant to hyper, can be explained as a different description of the same quasi-sensory experience 🫣
I've definitely not thought about dynamics as much but I think as long as people know (which they probably intuitively do) that a flickering image on-screen, for e.g., is just a proxy for whatever they consider "on" and "off" imagery, and not literally a replication, it seems reasonable to me!
If it's the same the image at the top of this thread, then my initial criticisms hold. If not, I'll check it out, and you can compare your experience to the quasi-sensory view too.
"Imagery can be compared to other types of experiences" Which experiences are reference experiences for what a mental image actually is? We need to define them.
Exactly. Everything follows from this. All claims about atypicality are impossible without an accurate reference for typicality. Hence the attempt to explain what quasi-sensory experience even is osf.io/preprints/ps...
I love phenomenological reports, and collect them too, but I just can't see how any would get past this argument which we've discussed before (originally from psycnet.apa.org/record/1989-...)
If imagery can be described as either seen or unseen, cannot be replicated with any real image, is unlike all other experiences, and every attempt to explain its appearance has failed, how can we claim we have validly measure atypical experience when we can't even describe a typical one?
If we ask all of them to visualise a black square instead, the idea that we can replicate mental image experiences with real images breaks.
The diff between mental and real images can't be reduced to blur/saturation/etc philarchive.org/rec/AMYIV philpapers.org/rec/BOUWAS
Researcher: Let's throw a ball on the weekend
P1: *brings a catcher’s mitt*
P2: *brings a ballgown*
Me: Shouldn’t we clarify what we mean by “ball”?
Researcher: No need, we have shown meaningful differences in how P1/P2 experience the world
Is this where we're at for ind diffs in mental imagery?
Also, I'm fuzzy on associator/projector stuff, but it could be a separate axis from non-sensory->quasi->sensory axis. The latter deals with how something is experienced in the vis. field regardless of depth info/environmental integration (i.e. on a surface, floating in space, or in/behind our head)
I think any real-image demo cannot mimic imagery - each pixel of a blurry/desaturated apple is still visually experienced. If a mental image truly matched those apple pics, there would be no imagery debate: it could be exactly replicated with a real image (a blurry real image is still a real image).