Summary of Wednesday's post

The other day I posted an exhaustive table of cognates between Old Icelandic and other Germanic languages in words that begin with <a>. It was to show how the comparative method can be effectively used, both to draw conclusions about the pre-history of the Germanic languages. It was a bit messy, however, here are the general rules we found in a cleaned-up format:

Icelandic and Gothic

In all instances, Old Icelandic a-initial corresponds with Gothic a-initial.

Icelandic and Old Saxon

In all instances, Old Icelandic a-initial corresponds with Old Saxon a-initial except...

  1. ...when the second syllable in OS was *i, in which case a-initial rises to an e-initial.
  2. ...Old Icelandic diphthong <au> corresponds with OS <ō>.

Exception 1: OIc. annarr "other" ~ OS ōthar, āthar, andar. Andar is the expected form. The nasal -n- was probably lost and the a- lengthened and nasalized to ā-. Old English raised and rounded  nasalized ā- to ō-, this may have unexpectedly happened in OS ōthar as well.

Exception 2: OIc. aptann "evening" ~ OS āƀand. We can speculate that a pre-OS *aft- reduced to aƀ- and a underwent compensatory lengthening.

Icelandic and Old High German

In all instances, Old Icelandic a-initial corresponds with Old High German a-initial except...

  1. ...when the second syllable in OHG was *i, in which case a-initial rises to an e-initial.
  2. ...OIc. diphthong <au> corresponds with OHG <ō> when followed by a front consonant and <ou> when followed by a back consonant (e.g. OIc. auð- "easy" ~ OHG ōthi // OIc. auka "augment" ~  OHG ouhhōn).

Exception 1: OIc. arfr "inheritance" ~ OHG ābund, āband. Evidently *-r- was dropped in OHG and the a-initial was lengthened to ā- as compensation.

Exception 2: OIc. at "at" ~ OHG az, ez, iz. First form is expected, latter two are not. Probably variation as a weak vowel (mirrored in OE and OF of).

Icelandic and Old Dutch

In all instances, Old Icelandic a-initial corresponds with Old Dutch a-initial except...

  1. ...OIc. diphthong <au> corresponds with ODu. <ō> in all instances.

Exception: OIc. aptann "evening" ~ ODu. āvont. See Exception 1 under OS.

Icelandic and Old Frisian

In all instances, Old Icelandic a-initial corresponds with Old Frisian e-initial except...

  1. OIc. diphthong <au> corresponds with OF <ā> in all instances.
    1. Exception: OIc. auk "besides" ~ OF āk, ēk. Second form is unexpected.

Exception: OIc. aptann "evening" ~ OF ēwend, iōwen. Second form is surprising.

Icelandic and Old English

In all instances, Old Icelandic a-initial corresponds with Old English a-initial or æ-intial (We cannot devise a rule for which one to expect on our data alone), except...

  1. ...when the second syllable in OE was *i, in which case a-initial rises to an e-initial.
  2. ...when OIc. a- is followed by <r> or <l>, it corresponds with OE <ea> or <ie> in all instances, with no way to predict which.
    1. Exception: OIc. allr "all" ~ OE ealǣl. First form in expected, second form is unexpected and possibly a spelling variant.
  3. ...OIc. diphthong <au> corresponds with OE <ēa> or <īe> in all instances, with no way to predict which.
    1. Exception: OIc. aurr "ear of corn" ~ OE ēar, æhher, eher. The latter two are unexpected. We assumed they were spelling variants, scribes attempting to demonstrate the long-e + a combination.

Exception 1: OIc. aptann "evening" ~ OE ǣfen. See Exception 2 under OS.

Exception 2: OIc. af "from" ~ OE of. This unexpected form is probably because the vowel is weak. Also in OF and a similar change in OHG (Exception 2).

Exception 3a: OIc. annarr "other" ~ OE ōðer. See Exception 1 under OS.

Exception 3b: OIc. anda "breathe" ~ OE ēðian. Because the vowel in the second syllable is -i-, *an- rose to *en- before the loss of *-n- and undergoing compensatory lengthening to ē-. Thus the word did not undergo the expected *an- > *ā- > ō- change.