11 Comments
User's avatar
Julio Nicanor's avatar

Excellent explanation of Predictive Processing!

It’s so good I’m going to have a hard time explaining it differently in my upcoming post! Thanks

Douglas's avatar

Timely! I am presenting a similar topic at a conference this weekend (afterwards I plan to share my abstract here on substack). What I find most interesting is the idea of manipulating one's 'network of priors' or 'web of beliefs' as a means to introduce new information instead of the status quo of 'persuading' or 'convincing' the mind's preexisting framework. That unique (and scary), and brings with it a host of new ethical quesitons

AffectiveMedicine's avatar

Very true. Would love to see your abstract when it's ready!

Michal Patarák's avatar

I don’t think serotonin can be easily conceptualized as the counterpart of dopamine – they are simply neurotransmitters with different functions that sometimes appear to stand in opposition, but that is only one perspective on their interplay. Likewise, I don’t think it is possible to simply define the role of serotonin in brain processes, in mental processes (which are, of course, distinct from purely “neural” ones), or in mental disorders and their treatment, given the diversity of serotonin receptors and the regulatory systems they are involved in. Nevertheless, I find your articles exquisite and enriching.

AffectiveMedicine's avatar

Thank you, I agree that reality is much more complex, of course. I guess that it bothers me that we do not have a "story" for serotonin in the way we do other neurotransmitters. I think that psychiatry's aversion to theory (perhaps due to our history with psychoanalysis) is a problem, and that we need theories (which are then falsified) to move forward.

Michal Patarák's avatar

Yes, I really like theories, but in the end I’m often disappointed by them — though that’s part of our field!

Siggi Prendergast's avatar

I don't think predictive processing makes psychedelics analogous to serotonergic antidepressants. In one case you have pretty selective agonism of the 5-HT2a receptor in Layer 5 of the cortex, in the other you have physiological agonism of all serotonergic receptors (most of which have conflicting effects on arousal).

Mirtazepine for instance, opposes the action of serotonin at the 5-HT2a (opposite psychedelics), 5-HT2c, and 5-HT3 receptors serving as either an antagonist or inverse agonist. Despite opposing serotonin, it produces the familiar side effects of emotional blunting at similar rates. Mirtazepine in this case *opposes* the plasticity inducing action that psychedelics promote.

Vortioxetine is another interesting case, its a both a serotonin reuptake inhibitor and receptor modulator with agonism at 5-HT1a, 5-HT1b, and antagonism at 5-HT3. Unlike Mirtazepine or other SSRIs, it is known in the literature to actually reduce emotional blunting while retaining antidepressant effects.

Food for thought, I do like the general thesis that the network actions of serotonin are inhibitory, but I think it runs into issues when considering classes of serotonergic drugs beyond SSRIs.

AffectiveMedicine's avatar

To be honest, I am pushing the frontier of my neuroscience knowledge. What do you think of arguments like in Opponency Revisited (Boureau and Dayan, 2010) https://doi.org/10.1038/npp.2010.151

Thanks for commenting!

Siggi Prendergast's avatar

Opponency doesn't strike me as an obviously good model, especially given the chart depicting equal opposing and collaborative actions of dopamine and serotonin. This is sort of the macro-level stuff that I think is better served by ascribing functions to brain regions as rather than neurotransmitters.

Siggi Prendergast's avatar

Really good and clear explanation of predictive processing by the way, last time I tried to put it into writing I sort of flailed.

AffectiveMedicine's avatar

Thanks! It's remarkable how confusing every academic paper on the topic is, not to mention books for a lay (really?) audience like Surfing Uncertainty.