SciPapers : Generational inheritance of memory in an animal model.
10
100Ṁ2409resolved Feb 19
Resolved
NO1H
6H
1D
1W
1M
ALL
I'm experimenting with using prediction markets to find interesting new knowledge, or (conversely) to show that some kind of knowledge is difficult to find. Market resolves to YES if we find the knowledge I seek, NO otherwise. Note: I will ONLY consider papers mentioned in the comments as eligible for the purposes of this question!
Still experimenting with the precise format, but here's my first shot.
Summary Description:
I'm looking for a paper that shows convincing evidence that a long-term memory can be inherited across four generations (parent to great-grandchild). That is, a certain animal learns a long-term memory of some kind, and then its great-grandchildren also have that memory even without experiencing the training that their ancestors went through. I want something that in my judgment is a "real" long-term memory (and not a kind of injury), and it want it to persist even if there's no way for the parents to communicate with their offspring.
Examples:
(note, a HIT here means that if you found a paper that shows this example, then it would cause the market to resolve to YES. A NEAR MISS means that if you found a paper with this example it would not, in my judgment, cause the market to resolve to YES, but only because it's missing a SINGLE salient aspect. If it had that aspect, then the market would resolve to YES).
NEAR MISS #1 ("famine"): A hypothetical paper notes that, If you expose a certain animal to famine, then its children will have epigenetic modifications that enable them to better survive famine, and tend more towards males. These changes persist for four generations even if the subsequent generations are not exposed to famine. This is a NEAR MISS because I don't consider it a "real" memory, but rather a type of injury. I consider the "training" in this case to be too limited, in that you can only choose "famine" or "not famine". If instead there were 20 different types of famine you could expose the animal to, and they each caused different changes in behavior, and these changes persisted over several generations in the absence of that famine, then I would instead consider this a HIT.
NEAR MISS #2 ("genetic mods"): A hypothetical paper states that you can modify DNA of an animal as an embryo, and induce a vast amount of heritable mental effects such as being more fearful in general, or avoiding a certain smell, or whatever. Although technically this "training" gets around the "famine" example, I still consider it a NEAR MISS because I feel that directly editing DNA is cheating. I want the memories to be possible to acquire, in principal, through direct sensory experience. If instead the paper showed that if you show the animal a picture that causes it to edit its OWN DNA, and that causes a change in behavior that lasts generations, then this would instead be a HIT.
NEAR MISS #3 ("teaching"): You point out that human children tend to speak the same language as their parents. This is a MISS because the parents are able to communicate with their children and teach them the language. I want a memory that still survives even without a direct line of communication between parent and child. If it was instead generally the case that children could start speaking their great-grandfather's language at age 13, even if up to then they had never heard a word of the language or seen their biological great-grandfather, then this example would be a HIT.
NEAR MISS #4 ("generational depth"): This paper: "Parental olfactory experience influences behavior and neural structure in subsequent generations", https://www.nature.com/articles/nn.3594 is close, but only shows learning from parents to grandchildren, not parents to great-grandchildren.
NEAR MISS #5 ("not behavioral"): This paper, "Transgenerational transmission of environmental information in C. elegans", https://www.science.org/doi/10.1126/science.aah6412 , shows that you can make transgenic worms that have two "glowy" states, "dim" and "bright", and that you can make them switch from "dim" to "bright" by changing the temperature, and that this change will last over 14 generations. I don't consider this a "hit" because the behavior of the worms doesn't change depending on whether they are "dim" or "bright". Also, it additionally fails as in the "famine" case, because "dim" and "bright" are the only two options. If instead there were 20 different "brightness" options and they each led to different behaviors, then this would be a HIT.
HIT #1: You find a paper with the following: Researchers trained three successive generations of rats to associate a particular room with foot shocks. Then the researchers separated the next litter of rats from their parents and for two generations did not train with foot shocks. Then the third generation of rats still exhibited freezing (a fear response) when placed in the same room their great-grandparents were in.
HIT #2: You find a paper w/ the following: Researchers trained mice to solve a certain maze. Then they raised three generations of those mice without any maze training. Nevertheless, the great-grandchildren of the maze-trained mice were able to solve the same maze much faster than control mice.
HIT #3: A crow learns, with some effort, how to eat a certain nut efficiently. It's eggs are raised by a different naive crow, and the resulting children and their children never see that nut in their lives. Nevertheless, the third generation of crows are able to eat the nut with similar skill as their great grandparent, without any initial effort to learn how to eat the nut.
HIT #4: Same as HIT #3 but the crow learns that eating the nut makes it sick and will avoid nuts that look like it from then on. It's great grandchildren still avoid the nut the first time they see it even if they've never before experienced the nut, and even though naive crows DO attempt to eat the nut.
----------------------------------------------------
How this Market Works:
1. If you know of a paper that meets my criteria, bet what you will and reference the paper in the comments. I'll check it out and include it in my market decision. You would presumably buy YES along with listing the paper.
2. If you think a paper that's already been listed clearly meets my criteria, you can do your own analysis of that paper and make your case in the comments along with buying YES. This helps me judge the paper and provide a fairer and faster market.
3. Even if you don't have a paper you can of course buy YES/NO based on whether or not you think the paper both EXISTS and that the community can FIND it in the allotted time.
4. If you have a paper that clearly shows that this type of thing is impossible, link that along with presumably buying NO.
Note: this is the first such question and I'm using it to calibrate and as a test. Full disclosure: for this market I already know a paper that DOES meet my criteria. If at the end of this market, no one has found a suitable paper, I promise not to swoop in and link to the paper that I already know would win. Of course, if anyone finds the exact paper I'm thinking of they will automatically win. It's my hope that you guys find something NEW that I don't know about. If no one finds anything worthwhile then I will have to resolve NO. Note again, I will ONLY CONSIDER PAPERS MENTIONED IN THE COMMENTS WHEN DECIDING THE RESOLUTION OF THIS MARKET.
This question is managed and resolved by Manifold.
Get
1,000 to start trading!
🏅 Top traders
# | Name | Total profit |
---|---|---|
1 | Ṁ32 | |
2 | Ṁ29 | |
3 | Ṁ23 | |
4 | Ṁ13 | |
5 | Ṁ9 |