Is there a pattern in the positions of long and short vowels in Old English?

Question

Long and short vowels aren't distinguished in writing in Old English. Are there any patterns I can use to guess the length of a vowel?

Answer 1

Here's a relevant answer by Alex B. to a prior question: What did we gain in return for the loss of phonemic vowel length from Old English?

That answer mentions that vowels in unstressed syllables are short as a rule in Old English. (This applies especially clearly to inflectional suffixes, which are unstressed in Old English as a rule. Compound words might have retained some amount of stress and therefore length on the second element. It is not always easy to distinguish between the use of derivational suffixes and compounding: derivational suffixes are likewise expected to be less stressed than the root they are attached to, but we apparently find an at least potentially long vowel in the second element of words such as freó-dóm "freedom", which suggests that the second syllable, although less stressed than the first, retained enough stress to support a long vowel.)

Alex B. writes that "Lass 1994 argues that vowel length was free in OE".

"Remarks on syllable quantity in late Old English and early Middle English", by Raymond Hickey (1986), likewise says

for monosyllables [...] vowel length was free: the structures VVC and VC occur abundantly in Old English.

(page 3)

However, there are some restrictions of "word minimality":

root syllables cannot consist of a single light syllable. The minimal requirement for a monosyllabic root is a heavy syllable: VV or VC.

("Prosodic Preferences: From Old English to Early Modern English", by Paula Fikkert, Elan B. Dresher, and Aditi Lahiri, page 128)

"A U-turn and its consequences for the history of final schwa in English", by Donka Minkova, clarifies that "vowel-final monosyllabic content words" always have a long vowel, but "The vowel in function words can be shortened and reduced" (Studies in Variation, Contacts and Change in English Volume 16: Can We Predict Linguistic Change?, edited by Christina Sanchez-Stockhammer, Friedrich-Alexander University Erlangen-Nürnberg, 2015).

You can in some cases use knowledge of modern English as a source of information, but that's pretty limited because of the many changes that affected vowel quality. But one correspondence that usually holds is that Old English a was long ā in words where it turned into modern English o(a) /oʊ/ (such as loaf, Old English hlāf).