Depression and Creativity
post by Bill Benzon (bill-benzon) · 2024-11-29T00:27:41.072Z · LW · GW · 0 commentsContents
Is there evidence for a relationship between depression and creativity? That’s tricky. A teacher of mine, the computational linguist David Hays, thought that depression might indicate a process of mental reorganization that precedes a period of creativity. It’s as though you get stuck on a line of thought and have to undergo something like mental hibernation. During tha... I think Hays primarily developed the idea from observing his own creative process. But he might also have been thinking about computers. Your remark about reorganizing data structures is suggestive. I know one thing that programmers do is refactor their code. That’s a specific kind of reorganization... Or think of an evolutionary process, where the species hits an evolutionary dead end. So it’s got to dedifferentiate and try again. Arthur Koestler used a French phrase to talk about this: “reculer pour mieux sauter.” It also relates to current work in machine learning and neural networks where an evolutionary style of computing is involved. Yes, it does. “Strategic regression” – I love it. None No comments
This is cross-posted from New Savanna.
I’m working on an article about depression and decided to ask about the connection between depression and creativity. Claude makes a few general remarks and asks for my thoughts. So I mention some remarks that my teacher, David Hays, had made years ago, mentioning that he was a computational linguist. After some general remarks, Claude asks:
... did Hays develop this idea through his work in computational linguistics? It seems like there might be interesting parallels between how computer systems need to reorganize data structures and how human minds might need to reorganize mental frameworks.
Interesting, most interesting. Our ensuing interaction felt a bit like a real conversation. I was surprised and pleased. This felt a bit like, you know, real intelligence.
I’ve appended its remarks below. I’m using my standard convention, setting my words in bold while Claude’s is plain.
Version: Claude 3.5 Sonnet
* * * * *
Is there evidence for a relationship between depression and creativity?
The relationship between depression and creativity is complex and has been studied extensively. The research shows several interesting patterns:
Historical observations have noted many renowned artists, writers, and musicians who experienced depression, like Virginia Woolf, Vincent van Gogh, and Sylvia Plath. However, we should be careful not to romanticize mental illness or assume correlation equals causation.
Scientific studies have found some evidence for links between depression and certain aspects of creativity:
Some research suggests that people with mood disorders may experience enhanced divergent thinking during certain phases, allowing them to generate more unique associations and ideas. The changes in emotional processing and perspective that can come with depression might contribute to seeing things in new ways or having deeper emotional insights.
However, severe depression typically impairs cognitive function and motivation, which can significantly reduce creative output. Many artists report that while their experiences with depression may inform their work, the actual creative process usually happens during periods of better mental health.
A key factor seems to be that creative pursuits can serve as a coping mechanism and form of emotional expression for people experiencing depression. The relationship may also be bidirectional - creative personalities might be more prone to rumination and emotional sensitivity.
What are your thoughts on this relationship between mental health and creativity? Have you observed any patterns in your own creative process?
That’s tricky. A teacher of mine, the computational linguist David Hays, thought that depression might indicate a process of mental reorganization that precedes a period of creativity. It’s as though you get stuck on a line of thought and have to undergo something like mental hibernation. During that “hibernation” the mind reorganizes.
That's a fascinating perspective from David Hays. His idea of depression as a period of mental reorganization or "hibernation" adds an interesting cognitive dimension to the discussion. It aligns with some modern theories about the potential adaptive functions of depression, particularly the "analytical rumination hypothesis" proposed by evolutionary psychologists.
This hypothesis suggests that depression's characteristic features - social withdrawal, reduced activity, and intense focus on problems - might serve to create mental space for working through complex problems or restructuring one's understanding. The "stuck" feeling you mentioned could represent this deep processing phase.
The metaphor of hibernation is particularly apt because it suggests both:
1. A necessary withdrawal or dormancy period
2. An underlying productive process that sets the stage for future growth
I'm curious - did Hays develop this idea through his work in computational linguistics? It seems like there might be interesting parallels between how computer systems need to reorganize data structures and how human minds might need to reorganize mental frameworks. His background in computational thinking might have given him unique insights into cognitive restructuring processes.
I think Hays primarily developed the idea from observing his own creative process. But he might also have been thinking about computers. Your remark about reorganizing data structures is suggestive. I know one thing that programmers do is refactor their code. That’s a specific kind of reorganization so it is more coherently organized.
The parallel between mental reorganization and code refactoring is quite illuminating. In refactoring, you don't change what the code does (its external behavior), but you restructure its internal organization to make it clearer, more efficient, or more maintainable. This often involves recognizing patterns that weren't obvious in the original implementation and finding a better way to express them.
This maps interestingly onto Hays's idea about depression and creativity. Just as a programmer might hit a point where the code has become too unwieldy and needs refactoring, a person might reach a point where their current mental models or ways of thinking have become inadequate for handling their experiences or insights. The depressive period could be like the uncomfortable but necessary process of pulling apart the existing structure to create something more robust.
What's particularly interesting is that both processes often require a kind of stepping back. When refactoring code, you often need to stop adding new features temporarily and focus on understanding the deeper patterns. Similarly, depression often involves a withdrawal from normal activities, which could serve a similar purpose of creating space for this deeper reorganization.
Did Hays discuss what might trigger these periods of necessary reorganization? I'm wondering if there are particular types of cognitive challenges or experiences that tend to precipitate this process.
Or think of an evolutionary process, where the species hits an evolutionary dead end. So it’s got to dedifferentiate and try again. Arthur Koestler used a French phrase to talk about this: “reculer pour mieux sauter.”
That's a brilliant connection! “Reculer pour mieux sauter” - step back to better jump forward - captures this idea perfectly. The evolutionary parallel really illuminates both Hays's theory about depression and the whole concept of creative restructuring.
In evolution, when a species hits a dead end, it can't just make small incremental changes anymore - it needs a more fundamental reorganization. Some evolutionary biologists talk about "despecialization" where organisms lose specialized features before developing new ones. This is similar to how stem cells need to dedifferentiate (become less specialized) before they can develop into new cell types.
This maps beautifully onto both:
- The depression-creativity cycle Hays described, where the mind needs to "dedifferentiate" or break down existing patterns before building new ones
- The creative process more generally, where sometimes you have to abandon your current approach entirely and return to a more basic state before finding a new path forward
It also suggests why this process can feel so uncomfortable or even frightening - you're temporarily giving up established structures or competencies in hopes of developing better ones. Just like an organism in the midst of evolutionary transition would be temporarily less well-adapted.
The French phrase really captures the intentionality or purposefulness of this retreat - it's not just stepping back, but stepping back specifically to gather force for a better leap forward.
It also relates to current work in machine learning and neural networks where an evolutionary style of computing is involved.
Yes! This connection to machine learning is really insightful. In artificial neural networks, there are several phenomena that mirror this "step back to leap forward" pattern:
- In gradient descent optimization, sometimes the system needs to move "uphill" temporarily (accepting worse performance) to escape a local minimum and find a better global solution. This is why techniques like simulated annealing deliberately introduce random noise or "temperature" to help the system explore more widely.
- In training neural networks, processes like "dropout" temporarily deactivate some neurons during training. This forced dedifferentiation helps prevent overfitting and leads to more robust learning - similar to how depression might force a "dropping out" of established mental patterns.
- In evolutionary algorithms and genetic programming, the concept of "reversion" or returning to less specialized states is built into operations like mutation and crossover. The system sometimes needs to break down existing solutions to explore new evolutionary pathways.
It's fascinating how these computational processes mirror biological evolution, creative thinking, and potentially the depression-creativity cycle Hays described. All seem to involve periods where existing structures or patterns need to be destabilized or broken down to enable new forms of organization to emerge.
Does this computational perspective add anything to your understanding of Hays's original insights about depression and creativity?
Yes, it does.
This feels like quite a profound convergence - we have parallel patterns across biological evolution, mental processes, and machine learning, all suggesting that progress sometimes requires a kind of productive regression or dedifferentiation. These aren't just metaphorical similarities but seem to point to a deeper principle about how complex systems develop and adapt.
The computational angle particularly strengthens Hays's insight by showing how this pattern emerges even in artificial systems designed purely for problem-solving. It suggests this isn't just a quirk of human psychology but potentially a fundamental principle of learning and adaptation in complex systems.
It makes me wonder what other domains might show this same pattern of "reculer pour mieux sauter." Are there examples in social systems, organizational change, or other areas where progress requires this kind of strategic regression?
“Strategic regression” – I love it.
What a great phrase to crystallize this idea! It really captures both the intentionality and the paradox - that sometimes moving backward is actually the most productive way forward. It helps remove the stigma from these apparent "regression" periods by highlighting their strategic value.
I imagine this concept could be really valuable in fields like education, therapy, or organizational development. It suggests that periods of apparent setback or loss of function might actually be necessary phases in development rather than just failures to progress.
This whole thread from Hays's original insight about depression through evolution, computing, and now to this broader principle of strategic regression feels like it's outlining something fundamental about how complex systems develop. Would you say this pattern of insight development itself might be an example of the principle - where we had to step back from the specific case of depression to see the broader pattern?
0 comments
Comments sorted by top scores.