Take the 2-minute tour ×
Cognitive Sciences Stack Exchange is a question and answer site for practitioners, researchers, and students in cognitive science, psychology, neuroscience, and psychiatry. It's 100% free, no registration required.

Several years ago, my adviser wrote an article discussing the observation that a hundred years of studying memory had not resulted in the discovery of a "law" of memory. He wrote (p. 247):

When one reads or hears the claim that “repetition improves retention” or that “generating information enhances recall” or that “forgetting follows a power function” or all other statements of the form that “variable X has a consistent effect on variable Y,” one can be certain that the claim is both true and false. It is true in that conditions can be found under which the rule holds (otherwise the claim would not be made), but false in that a skeptic can always say: “Very nice work, but your finding depends on many other conditions. Change those and your effect will go away.”

Has the story changed since 2008? Can anyone cite a proven true law of memory? If not, what's the closest thing memory has? The idea of transfer-appropriate processing is the closest I can name, is there something more substantial?

share|improve this question
Psychology really isn't a field where "universal laws" are used or evaluated. I think you're looking for something that, more generally, doesn't exist. I've never heard "laws of memory" used as a phrase or thing except in that one paper. The word law itself is certainly misleading if not a strawman. –  Ben Brocka Feb 16 '12 at 21:01
That's a good answer! Although why wouldn't psychologists search for laws of memory? My Google Scholar search shows that the topic has definitely gotten some attention... –  Andy DeSoto Feb 16 '12 at 21:23
Much of the problem seems to be defining "law" here. One of those articles calls Gestalt Psychology a "law". Very few if any of the papers there could be considered as trying to find or propose a "law" as Roediger attempts to define it. Psychology works upon theories, which may or may not be generalized. "Law" is occasionally used to point to a specific apparently causal relationship but is not a formal, agreed upon thing in psych, nor is there a collection of "laws". –  Ben Brocka Feb 16 '12 at 21:32
Formally a law is an axiom of a theory. Laws of a theory allows you to derive theorems, and defines its level of soundness, completeness and falsifiability. Every theory has laws. Good laws make good theories. Human memory, contrary to the initial beliefs in the domain, is highly integrated to other (more complex) systems, such as consciousness, reward systems and learning; and this should be the simple reason why we don't yet have a good theory of it. –  Kemal Taşkın Feb 17 '12 at 7:11

2 Answers 2

up vote 2 down vote accepted

Memory is only a psychological construct. In reality there is only the brain, and the physiological functions of the brain. "Memory" is just a way to conceptualize some of these functions. For this reason, it's unlikely that a law will ever emerge. Part of the reason is that memory can refer to so many different things, biologically.

Even when you get down to the actual biology, laws are not really the standard. Biology is a messy thing and virtually no laws exist that couldn't be contradicted. Something as basic as the cell theory could be reasonably argued against. Viruses are not cellular, yet an argument could be made for them being alive. It isn't the current thinking, but it's not an absurd proposition either. If psychological phenomena are a result of biology, and no laws exist in biology, I think it's reasonable to say no laws will exist in psychology either.

Perhaps one of the strongest ideas about memory is Hebbian learning. But even that is just a simplification.

share|improve this answer
a dynamic system can have a laws at a higher level of description, without relying on laws at a lower level of description. Common examples are laws of thermodynamics (which abstract away almost all the underlying physical nuances. Although we can get them back through statmech, but that is not necessary) or general laws of evolution (not necessarily instantiated through biological evolution). –  Artem Kaznatcheev Feb 16 '12 at 19:38
I hadn't thought of it that way, Preece, thanks for the thoughts! Artem, that's a good counterargument too. Smart all around. –  Andy DeSoto Feb 16 '12 at 20:33
You make a good point. I'd considered evolution, but I really don't think there are any laws, per se. Nothing absolute can really be said about it, there are multiple mechanisms, multiple criteria. You're right that a dynamic system can have emergent laws. But I think that is only true so much as the underlying system is regular. Consider trying to find laws pertaining to a synapse. The laws will only hold to be true to the synapses you've collected data from. There could be a synapse that works in the exact opposite way. In fact, they have found such a thing! Anti-hebbian synapses. –  Preece Feb 16 '12 at 20:37

The article you cite does a wonderful job of tearing down an excellently crafted strawman. It is true that it is very difficult to meet his extremely precise measure of a "law". The two problems are Who is proposing "laws"? and Why are these criteria fair?

I have almost never heard of anyone in any psychological field posit that their theories are scientific laws. I have never seen an article so boldly state "X will cause Y in all possible cases", the brain is simply too complex and this is why research always includes information about it's sample, measure and test procedures.

A lot of effort is put into generalizing theories but attempting to generalize to the level of an absolute law is impractical and often unhelpful; maybe repetition doesn't help retention of knowledge when the participant is on fire. This is why the generalizability of results are always investigated and why no sane, accredited psychologist attempts to say their theories hold in all possible situations.

Further, the guidelines Roediger, H.L. (2008) set, specifically expecting a "Law" to stand up to variance in all factors of Jenkins’ (1979) "tetrahedral model" are compeltely unmeetable for a number of reasons.

One of the primary problems with his expectations is they require a law to generalize to images, episodic events, written words, ect. There are different types of memory. Episodic memory, Semantic Memory, Procedural Memory, working/short term memory vs Long Term memory, different types of memory function differently and operate in different parts of the brain. Memory loss affects different types of memory.

Anterograde and Retrograde Amnesia for example generally don't affect Procedural memory but mostly Episodic Memory.

Bottom line, Roediger, H.L. (2008) expects The Law of Apples to apply to Oranges. The reason no Law of All memories exists because different types of memory are fundamentally different, and any attempt to generalize something to all possible types of memory would be absurd.

He is right in so far as the field of memory is complex and universal laws don't exist for it, but no one active in the field seems to be making these assumptions. If you are looknig for these laws you are simply looking for the wrong thing, and should rather learn more basic things about memory, like the types of memory, types of learning ect, before you continue any investigation into the field.

share|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.