Grabbing ahold of languageWhy is "ahold" not like "akin"?
The prefix "a-" in English has a couple of major uses. One is that it is a negativising prefix, borrowed from the Greek. So "amoral" means "without morals". Compare with "immoral", which uses the Latin negativiser "in", which assimilates to a succeeding consonant in English.
The other is to give the sense of a locative adverb, in other words, something "in" or "at". So we say "ahead" because something that is "ahead" is "at the head", or in front; we say "abreast" to say that something is "at the breast", or next to. These words are adverbs because they modify verbs. You "look ahead" or "come abreast".
Interestingly, the word "adown" used to exist in English. It meant, simply, "up". But it has fallen into disuse, presumably because it is ambiguous which of these prefixes was desired.
It's pretty clear that "abreast" and "a breast", and "ahead" and "a head" are different constructions. The first is an adverb, the second an article plus a noun. This is because "a" is, as well as a prefix, the indefinite article. Of course, in speech, they sound the same. But if you write "He picked ahead" when you mean he picked a head of cauliflower, you are making an error in the written language. Similarly, "the ships steamed a breast" means something entirely different from your intention if you mean to say they were side by side.
To write "the ships steamed a breast" would in this case be a clear mistake, and we would consider someone who wrote it believing it to be correct to be illiterate in English.
If I write "grab ahold of my hand", I'm equally guilty of illiteracy. I do not intend to use an adverb and no sense can be made of the sentence as it stands. I mean to use the noun "hold" with an indefinite article. This is not disputable in English because in fact "ahold" does not mean anything in any sentence, and is always used in contexts where "a hold" is correct.
"Akin" actually adverbialises an adjective, similarly to "alike" and "around", although not "about", which has a different etymology. You do not say you are "a kin" to someone. You say you are "kin" to them. There is a noun "kin" but it cannot be used with an indefinite article because it is a mass noun.
It's true that people write "ahold" in "grab ahold of my hand". This is because they are not sufficiently literate to analyse "a hold" correctly in the spoken language. This kind of literacy does not have to be taught explicitly.
Competence in written language has various components. One is the ability to transcribe the spoken language correctly. This is a quite separate issue from the ability to write correctly in a register, which some take to be the meaning of "literacy". While we might in the common parlance consider someone to be "literate" only if they can "write fancy", functional literacy is much broader.
Why must we insist that people write correctly? Why can't they write what they like?
The bottom line, which may not be traded away, is that language is transactional. We trade symbols that have negotiated values. We agree, in writing English, that what we are writing is English. We agree to abide by the negotiated values of standard English, unless we are writing in some format that has different values, such as a text message. We don't agree to this because we like following rules. We agree because we like to be understood. When English was not standardised, it was difficult to communicate because each reader had to try to figure out what each writer was actually trying to say. All languages tend to a written standard simply because that makes them a better medium for communication than simply trying to transact with our own private renderings of the spoken language.
When we talk about what is "correct" in a language we are referring to the utterance in the context of where in a hierarchy of lects it belongs. We each have an idolect. This is the language we personally speak. It differs from person to person. Where it coincides with the idiolect of others, together we speak a dialect. Dialects can form a greater whole, which we know as a language. There's no hard and fast dividing line between dialects and languages, at base because there is a conflict between functional definitions (we speak the same language as other people who understand what we are saying) and social definitions (we speak the same language as other people who are either part of our ethnos or understand themselves to speak the same language as our ethnos). This leads to inconsistencies such as languages that contain dialects that are not mutually intelligible, and languages that are mutually intelligible to a degree that some dialects are not. For instance, on one reading one could consider Danish and Bokmal Norwegian the same "language" but their speakers insist they are not.
However, in written language, your idiolect tends to be a function of how well you can write the standard language. The picture is muddied by our use of different registers, just as it is in spoken language, but this is certainly truer of writing than it is of speaking. In other words, you build your written idiolect from the standard down, whereas spoken languages are built from the idiolect up.
Conventionalisation of written language obscures that the process is the same in both written and spoken language. One can use any words one likes in both, but one cannot enforce one's words on others. The listener or reader must agree that what we are using are in fact words. Because the point of language is to communicate our meaning, should the recipient of our linguistic act not agree that it is well formed, we are faced with our intent's being thwarted. In this, you cannot be an individual, however much you prize your individuality. Language is in nearly all cases not something you mutter to yourself. (Interestingly, you cannot even mutter intelligibly to yourself unless you use well-formed words, nor can you write a diary that you yourself can read unless you are certain that your future self will agree with you on what the words in your private language say.)