a word from Vocabulary Dutch
For analyzability, I have concentrated on contemporary language use: do modern speakers consider a certain form to be a derivation/compound or not? This may involve, for instance, the word draad ‘thread’ being labelled unanalyzable, although from a historical point of view it is derived from the verb draaien ‘to turn’ - the suffix -d has, however, for many centuries been unproductive. Likewise, the old and by now unrecognizable suffix –el is regarded as being unanalyzable; it was formerly used to form instrument names such as beitel ‘chisel’, schoffel ‘hoe’ and diminutives like eikel ‘acorn’. And sprinkhaan ‘grasshopper’, finally, is regarded as semi-analyzable from a contemporary point of view: it has haan ‘cock’ for its second element, obviously, but the first part - the stem of the verb springen ‘to leap’ – is no longer recognizable as such.
On the other hand, hangmat ‘hammock’ is counted as an analyzable compound: the word was borrowed from Spanish hamaca, but through folk etymology it has been adapted in such a way that modern speakers regard it as being a compound of hangen ‘to hang’ and mat ‘mat’.
For the (un-)analyzability of verbs, the following arguments have been used: Dutch has a small group of verbs in -n such as slaan ‘to strike, hit, beat’, doen ‘to do’. The great majority of verbs, however, end in -en – these are all considered to be analyzable, even when they have been borrowed. They are either semi-analyzable (like sterven, the first element having no meaning), or analyzable derived (like hoesten: cough-INF, see hereafter), or analyzable compound ( aansteken: aan-steken [on-stick-INF]. If a stem has both a noun and a verb, like hoest ‘cough’ and hoesten ‘to cough’ or slaap ‘sleep’ and slapen ‘to sleep’, it is, within our contemporary linguistic system, impossible to judge which of the two is the original form and which the derivation – even historically, it is often not feasible to determine this. In these cases I have consistently labelled the noun (slaap) unanalyzable and the verb (slapen) analyzable derived ([sleep etc.-INF]). Dutch has a number of prefixes for verb derivation that are very old and whose meaning cannot clearly be described: be-, ver- and ont-; for these prefixes I have usually indicated only whether the verb derived is transitive or intransitive.
My dating of the words is largely based on the information provided by the EWB and van der Sijs 2001. Most of the dates mentioned in these works come from the quotations given in the three big historical dictionaries of the Dutch language, the WNT, the MNW and the VMNW. For the period up to the 13th century we have few sources, and the dates for this period have always been found quite by accident. We know about many native words that they must already have existed then as well, but it is not until the 13th century that we have a reasonable number of texts at our disposal. But also dates found for other periods are always more or less relative and accidental. For French and English there is a rich tradition in the domain of datings, but even there scholars keep on correcting and supplementing. In Holland, the situation is much less propitious - datings have been investigated and provided for only a few decades there. Not a single date can be accepted as absolute, every date is an indication for a period. And it is "work in progress": more and more sources have of late become available digitally, making it possible for ever better and older locations to be discovered. Still, the data I have supplied give an indication about the century since when the word has been known in Dutch.