Like many other common practices in modern society, learning a whole new language as an adult is an unnatural phenomenon. Humans are programmed to acquire language, but they’re programmed to do so during childhood. While it’s perfectly natural to constantly learn new words, pick up novel ways of inflecting our voice, and so forth, there’s nothing natural about going from zero to fluent in a language starting when you’re 25. And like many other behaviors and conditions that were unknown to our hunter-gatherer ancestors, there are a lot of problems that occur when you try to learn a language as an adult.
One of the key differences between acquiring a language as a child and acquiring one as an adult is that the world of conversation within childhood is such that any given word that’s important will come up quite often, and the world of conversation within adulthood is such that many important words come up quite rarely. As a young child of 5, you’re learning concepts while you’re learning vocabulary; your world of perception is simpler than that of an adult’s, and as a result there’s a large swath of words which wouldn’t enter the picture, since they would be signs for concepts yet learned. Your days are still filled with fascination concerning relatively simple things, like foods that every adult in your culture is familiar with. As an adult learning a foreign language, on the other hand, the most compelling movies, interesting books, and fascinating conversations will make use of a much more varied world of vocabulary. With a smaller amount of vocabulary used in a child-like setting, any given word is bound to come up more often. And with a larger amount of vocabulary used in an adult-like setting, even important words might come up only every few months. A child bites into a juicy apple, and their mother uses the word “apple”; the following week they’re given an apple in their kindergarten class. An adult reads an article about an adopted child, and then that word (“adopted”) doesn’t come up again until 4 months later.
Of course there are plenty of words that children learn that don’t come up very often, but I think you may be able to see the overall point. The world of perception of a child is simpler, and thus the variation of vocabulary will be simpler. A thousand conversations between an adult and a child would certainly amount to a lot fewer unique words than a thousand conversations between the same adult and a peer of theirs. The world of communication in childhood includes a smaller subset of words, and thus each individual word will be likely to come up somewhat often. Adults, on the other hand, communicate about a larger variety of concepts (because they know about a larger variety of concepts), and thus the average word will come up less often.
Now, consider how memory works. When you learn a word, your brain stores the word for a certain amount of time. If you don’t encounter that word again within that amount of time, the word gets erased. And consider again that learning a whole new language as a child is the natural state of affairs, while doing so as an adult is evolutionarily novel; thus it wouldn’t be unreasonable to assume that the length that humans tend to remember new words within the early stages of acquisition meshes with the frequency by which such words tend to show up within those early stages. In other words, when acquiring one’s native language we may imagine that the memory interval of new words is high enough for the encounter interval of such words. If a child tends to remember new words for roughly a week, then the average word would come up at least once a week. On the other hand, since the world of communication in the adult world includes a larger variety of words, we can imagine that running the natural memory interval for new words acquired during the early stages wouldn’t work. You would tend to remember new words for a week, but many important words would come up only once every 3+ months. You’d often forget what you learn, and you’d get the sense that your mind isn’t soaking up the language the way it would work for a child.
In order to take the natural memory interval (i.e., the memory interval tuned to childhood acquisition) and make it work for learning a language as an adult, there are two major practices that have come about within the language-learning community: SRS and repetitive listening. With SRS, you USTAUP the encounter interval directly; even if you tend to remember words for only a week while many important words come up only every few months, putting the new words into an SRS ensures that those words are brought to your attention at a frequency that meshes properly with how long you tend to remember a given word. Repetitive listening, on the other hand, does something fundamentally the same; while a native might watch a YouTube video about someone who was adopted once and then not encounter the word “adoption” for another 3 months, a foreigner learning the language could listen to the video several times over the course of a few weeks, causing the subset of words that come up in the video itself to essentially appear more often than they would in natural communication.
But the problem isn’t just that adult learners at the early stages of the process of language acquisition tend to remember new words roughly as long as children do, but that many of the new words they learn come up less often than the new words that children learn, meaning they’re likely to forget what they learn unless they consolidate what they learn with SRS or a traditional technique of writing down new words and drilling them on a regular basis, or unless they do repetitive listening (or repetitive reading, for that matter). It’s also that adult learners tend to forget new words that they learn faster than children do. In other words, it’s not simply that the encounter interval of the average word in the world of adult communication far outstrips the default memory interval of the human mind for acquisition of language; there’s also the problem that the memory interval of the average adult learner’s brain is actually shorter than the default memory interval for a child learning a language.
The reason for this difference is related to the well-known phenomenon where children naturally pick up a native accent, and adults usually don’t. Basically, if you don’t have a clean grasp of the building blocks of the language, it’s harder to remember words. For instance, Japanese doesn’t differentiate between /m/, /n/, and /ng/ when they come at the end of a word, and thus Japanese students who haven’t burned that English-pattern distinction into their heads will have trouble putting “Kim”, “kin”, and “king” into different slots in their mind. And Japanese has a lot fewer vowels than English, so often a Japanese person will learn a word like “cut” and then a few minutes later they’ll say it like “cot”. When words shift on you, and you can’t create a clean slot in your mind for each new word, you will have trouble systematically tying words to meanings, which will make it feel like you have a bad memory. But really, it’s just that you need to spend a while drilling the fundamentals of the sound system of the language, so you can start to store the words in your memory properly.
The adult brain doesn’t construct a robust system of building blocks for a new language automatically, but that doesn’t mean you can’t do it manually. Most people jump straight into learning vocabulary and grammar, but of course babies spend several months learning to chunk the stream of sound that they’re hearing into a system of phonemes well before they start to learn how to wield words and syntax. Perhaps adults should start with that step as well, and then they wouldn’t feel so inferior to children when it comes to picking up language. Children aren’t spongy by dint of some sort of magical power of youth; there are multiple widespread psychological differences between a healthy child and the average adult learner, which can explain the famed power of children to absorb language like a sponge.
The conclusion is that while it’s generally true that children absorb language effortlessly, forgetting very little and soaking it all up like a sponge, while adults tend to go through a lot of painstaking study only for the house of cards in their mind to come crashing down if they decide to take a vacation from studying, we needn’t explain this as the result of an inescapable neurological difference between children and adults. We can simply point to the greater variation in the world of adult communication, necessitating SRS or repetitive listening, and then point to the lack of automatic tendency for adults to create a clean system of building blocks in their mind for their target language, which can be addressed by minimal-pair testing and other such techniques. While the encounter interval of the average word for an adult learner is longer than that for a young child (because of larger variation in the content), and the memory interval tends to be shorter (due to the same issues which cause a heavy foreign accent), it’s possible to deal with this issue, and restore the sponginess of a child, where you remember what you learn with no issues.