Following the example of 'take' -> 'taker, 'create' -> 'creater' and so on, we might expect that there would exist a noun '*raper' from the verb 'rape'. Instead we have the noun 'rapist'. Is there a term for such non-words that are formed according to the derivation rules of a language but don't exist?
To expand on snailboat's comment above, you might want to consult Laurie Bauer’s ‘English Word Formation’ (Cambridge UP 1983) for some background here. Quote (p. 87):
Blocking is the name given by Aronoff (1976: 43) to the phenomenon of the non-occurrence of a complex form because of the existence of another form. The form which causes the blocking may itself be complex or simplex. For example, Bolinger (1975: 109) points out that, despite the productivity of –er suffixation in English, there is no word ‘stealer’ because of the prior existence of the word ‘thief’ which carries the appropriate meaning; the existence of forms like ‘bad’ and ‘small’ blocks the formation of *ungood and *unbig; the prior existence of ‘enlist’ prevents the use of ‘list’ as a verb with that meaning (Clark and Clark, 1979: 798).
In fact, ‘stealer’ or ‘raper’ would be termed by Bauer ‘nonce formations’: ‘new complex words coined by a speaker/writer on the spur of the moment to cover some immediate need.’ (p. 45) Blocking takes place when these fail to lexicalize. As Bauer details, linguists have suggested that there are some suffixes whose formations are never blocked, for instance –ness. This leads them to posit ‘an inverse relationship between productivity and lexicalization’. It also implies that productivity is a cline phenomenon. Interesting stuff!
(Aronoff 1976 is ‘Word Formation in Generative Grammar’, by the way.)
I would say the closest you can get is "infantile speech", not because it actually is infantile speech but because I think people would understand what type of words the one you are referring to is.
The words you are referring to, like *goed, are obtained by overgeneralising rules, which is usually what children do and what adults try to correct. Actually, the *goed was one of the examples in a seminar child speech development I goed to ;)