Since 1994, on and off, I have occupied myself with the circumstance that some have memories, so early that they defy conventional memory psychology's theory of "infantile amnesia". In particular, memories from one's first year of life have had my attention, somewhat later memories apparently being relatively common, despite the usual claim that no one can remember anything from one's first three years of life, which is the period normally quoted regarding infantile amnesia.
While I did succeed in finding individuals reporting first-year memories, I never came across a (for me) satisfactory explanation as to why some have these memories and others do not. For that reason I have finally decided to formulate my own hypothesis. It is below.
First, a recapture of what is actually meant by "infantile amnesia": It has been observed, so one claims, that infants do store memories, but can not retain them as long as can adults. The younger the infant, the shorter the storage. Only beyond the age of three, it is said, begins the permanent storage of memories. A fairly recent hypothesis proposes that it is the high rate of formation of new neurons in the hippocampus during infancy that erases memories; the decreasing neurogenesis with age corresponds to (and causes) the increasing ability to retain memories, according to this plausible hypothesis. Relevant in this respect is also the fact that the brain is the only organ of which the cells are not constantly being replaced throughout one's life; if that were the case, one might never be able to retain memories very long, but be losing one's memory, and even part of one's personality, over and over again. Also, vital functions like heartbeat and breathing might stop, as those are controlled by the brain stem.
So, the question regarding early memories is, how do some memories of some individuals escape being erased in this period of high new-formation of neurons?
I suggest that those with early memories have, or had as infants, a style of thinking and storing information that is more robust against neurogenesis than is the style of most people. This robust style I would identify as abstract thinking and procedural information storage, as opposed to concrete thinking and declarative storage as practised by the majority. I suspect that the abstract, procedural style is to a lesser degree localized in the brain, more "holographic", involves more associations with already stored concepts, and therefore is less vulnerable; a memory thus stored is less likely to be erased by a local change like the addition of new neurons.
While there may be a concrete/abstract continuum more than a dichotomy, the styles are best explained by describing their extremes:
Although perhaps caricatural, the following images illustrate the nature of concrete thinking and declarative storage: Some twentieth-century composers employed a style of composing music called musique concrète; sounds, like nature or city sounds, were captured with a tape recorder and a collage was made of them, which was simply played back to the audience by way of performance. Something similar was done in visual art with so-called "ready-mades". One may also think of a cook who merely opens a few tin cans to prepare a meal; this latter image serves to illuminate only the retrieval aspect of memories, of course.
With some exaggeration, one may picture concrete thinkers as walking recorders who simply store perceived information more or less verbatim and play it back later. It is thinkable that memories in such individuals are relatively easily erased by local physical changes in the brain. Extremely concrete thinkers would be characterized by their abundant use of idiomatic expressions, clichés, and prejudices — which are the "tin cans" or "ready-mades" of language — and by their lack of originality and comprehension. Also by their random illogical mangling of what has priorly been communicated to them; since they do not abstract information before storing it, possible errors in storing and retrieval will be on the concrete level, so random and illogical. Text and other work produced by concrete thinkers will be bound to its specific period and culture, and quickly appear dated.
Abstract thinking and procedural storage are more analogous to traditional composing, where music is notated in symbols (so, abstracted) and interpreted by musicians. The actual music lies not only in the notation, but involves also the knowledge to read that notation, and the skill to play one's instrument. The eventual sounding composition is thus spread over multiple individuals and fields of expertise, rather than existing in verbatim storage on a single physical object like a magnetic tape, and will be harder to erase or destroy. In cooking terms, abstract thinkers are "cooking from the ingredients", the result depending on recipe and skill instead of on the stored contents of a can.
Abstract thinkers are those who abstract, conceptualize, associate, interconnect their perceptions before storing them. They store not so much the actual information but the procedure to obtain that information; they store the information's associations with other information, using methods like redirection. Thus, the information will be less localized, more global, holographic as it were, and harder to erase by a local physical change in the brain. Extremely abstract thinkers would be characterized by their literal use and understanding of language, their avoidance of clichés and idioms, their prejudice-blindness, their originality, their formation of new words, and their depth of understanding. When they make errors, these take place on the logical, conceptual level, quite unlike the random mangling one sees with concrete thinkers. Work by abstract thinkers tends to last, to transcend time and cultural barriers.
After forming the above hypothesis, a logical next step is to ask what causes the individual to incline to either abstract or concrete thinking. Interesting a question as that is, its perusal is not properly required in the present context but perhaps more suited for treatment as a topic in its own right, or in some other circumstance.