I have recently just realized that in English, sometimes the same word will have different part of speech depends on the way you pronounce it.
For example, record can be a noun or a verb depends on where you stress the pronunciation.
However, in many romance language (at least to my knowledge in French and Spanish) this never happens. Record (verb) would be "enregistrer" in French and record (noun) would be "enregistrement" in French.
Is there a special term to describe this phenomenon?
As you are looking for a term, it is zero derivation. English has zero derivation turning a verb into a noun and vice versa. Note that 'record and re'cord aren't an example for zero derivation, in the spoken language they are clearly different words. In Romance languages, part of speech is to a large degree determined by the word form and to derive a verb from a noun or vice versa, you need to change the word form, usually by adding some morphemes.
You're quite right; an English word can usually be used as a noun, verb, or adjective in some context (though which one varies), and most European languages don't work that way.
This is due to the fact that most European languages are far more inflected than English is. English is much more like Chinese, where the same sequence of sounds may mean many things; but English doesn't have characters to display them. Instead we use a sort of an alphabet, which sort of represents some parts of the pronunciation, if you can guess right. Also like Chinese.
In German, for instance, which is closely related to English, you can see what English might have been like if it had kept all its inflections. There is never any doubt about whether a word is a verb or a noun in German; either way, you rarely see the root word, only inflections and derivations from it. Verbs have two tenses, same as English, but every verb form indicates number, tense, and often person of the subject. This is only true in English of the verb be (if it is a verb).
Even German infinitives have a special suffix to mark them as infinitives. Plus, every verb has a separate set of subjunctive forms that mark almost all the same distinctions. This is not true in English of any verb at all.
As for nouns, there is only one plural suffix in English, and its use is quite regular; there is no gender or case, except in personal pronouns. In German every noun has its own plural form; there are about a dozen variants. Plus there are four cases and three genders (actually four, since the plural inflections all work the same for every gender). Usually the nominative is the root, but the gender and plural forms, and all the case forms, must be learned individually for every noun.
Adjectives are even more complex in German, and totally uninflected in English; I won't go into the details.
I would like to begin by arguing that record (as in record shop) and record (as in can you record that?) are not the same word. As you have noticed, they differ in pronunciation which goes back to which syllable their stress is on. As far as I am aware, a native speaker would always use the correct pronunciation when asked to pronounce one of these words with sufficient context and conversely, a native speaker would always be able to identify whether the utterance was the verb or the noun. This extends to their derivatives too: records (that one can buy in the shop) and records (the thing he did) retain that difference.
Naturally, the difference between the two words is not reflected in spelling – but neither is the difference between plane and plane (which have different etymologies, deriving from different languages). So spelling is not always the best indicator.
Almost all such cases that I am aware of are a pair of verb and noun that derive from the same root. I suspect that this difference arose in older forms of English when grammatical marking was more widespread. In Middle English, for example, verbs had endings depending on person and number that gradually folded into the mostly unmarked verbs of today that only distinguish 3rd person singular and past tense. I would assume that the stress difference originates here. If you assume an ending at the end of a verb, that would put stress consistently in penultimate position. After the endings were lost, the stress of verbs – while still on the same sound – would have moved to final position.
That this can happen in English in this way is a direct consequence of the loss of inflection English has had. By comparison, German has mostly retained endings and inflection and verb and noun typically differ in more than just one feature (e.g. Aufnahme versus aufnehmen; although cases such as Essen versus essen – pronounced identically – also exist). Given that the Scandinavian languages have also lost most of their inflection, such pairs might have turned up there too; I don't speak any Scandinavian language well enough to know.
However, that this happened was a result of chance and by bringing together some not-entirely-common developments in a relatively specific manner. Most of the world's languages either don't have the necessary prerequisites or did not undergo the necessary changes so they did not end up with such features.
Both French and English have zero derivation between different parts of speech.
In English, the conventional citation form of a verb is the "plain form", which also serves as the infinitive, and which ends in a "zero morpheme". (Sometimes to is added before it.) Therefore, there are many noun-verb pairs related by zero derivation in English that have identical citation forms (although for record the noun and verb are only identical in spelling, not in pronunciation).
However, if we happened to cite English verbs by their third-person singular forms, or by the gerund-participle form, while still citing nouns by their singular form, we'd get something different: the verb records (as in "it records") or recording (as in "I'm recording the show") vs. the noun record (as in "play the record").
French verbs are conventionally cited by their infinitive form. This form never ends in a zero morpheme in French: all French infinitives end in an infinitive marker of some type (conventionally, verbs are grouped by the following infinitive endings: -er, -ir, -re, -oir). If you cite a French noun-verb pair related by zero derivation, the verb's citation form will have one of these infinitive endings, most often -er, as in the verb blâmer and the noun blâme (the CNRTL says the noun is derived from the verb).
However, the third person singular form of French verbs often ends in a zero morpheme (or in some theories, an underlying schwa sound that ends up being silent). The verb form blâme used in elle blâme "she blames" has the same spelling and pronunciation as the noun blâme used in le blâme "the blame".
It's also possible in French for the infinitive form of some verbs to be used as a noun, as in un être humain "a human being".
