There are a number of reasons for variation in Old English spellings. People generally first think of regional variations in pronunciation. This does not only show up when comparing a text written by two different scribes, like Beowulf, where the second scribe spells his name Biowulf throughout (and generally replaces the <eo> diphthong with <io>.

If we are talking about a single scribe, however, it would take a detailed examination of each variation to come up with an explanation. Some common ones are that he is copying from a manuscript whose spellings reflect a particular regional dialect or time that differs from the current scribe’s. He may in one instance copy the manuscript reading exactly, while in another he may “update” the spelling to his own local and contemporary pronunciation. So you could have a late tenth-century West Saxon scribe copying a ninth-century Mercian exemplar, with perhaps a mixed output. You can really see this in texts where the scribe seems uncertain about the meaning of a particular word or passage he is copying and so he copies the exemplar very closely.

Another reason depends on the lack of clarity of which exact vowel is being said in a particular word. For example, the pronunciation of [a] and [o] before [n] seems to have been somewhat difficult to determine and so a text might show variation between land and lond or mann and monn, though other scribes are more consistent in their usage. Where this uncertainty really shows up is in the final or inflectional syllables. Is the scribe saying drifen or drifon or drifan? He might not pronounce them distinctly enough to know. The later we get in the Old English period the more of a problem this becomes.

Other explanations could be found. Some words seem to have had something like a set, traditional spelling, which a scribe could use some of the time, while switching to a less traditional spelling now and then, and so on. It’s a difficult question about whether or not these should be considered “mistakes.” The standard answer is clearly “no,” that they had a very different attitude towards writing than we do, and I would support that answer. On the other hand, there are words that seem to have consistent spelling preferences and words that were preferred over others. While using a different spelling or word would not really be a mistake, it might be a sign of a moment of inattention by a scribe, in the same way that lapsing in the updating of a text’s spelling could be more due to inattention than any other reason. That still doesn’t mean it was a mistake, however, but there are plenty of cases of correctors (sometimes the original scribe and sometimes a different one) going through and fixing spelling or grammar.