the suffix *-́

i̯ă

-/-

i̯ā́

– was leveled to *-

i̯ā́

-an original combination *-

u̯i-i̯eə₂

– survived as *-

u̯i̯ā́

-.

From the evidence of Linear B texts, we see that

palaió

– ‘not new’ is probably a thematization of

pálai

‘near in past time’ (see Chadwick 1976). Perhaps

krataió

– is likewise from *

kratai

plus –

o

-? But

kratai

– is not attested as an adverb like

pálai

. Or perhaps

krataió

– is *

krata

plus –

ii̯ó

– (cf. adverb

kárta

)? But how to explain the accent of –

ii̯ó

-?A reconstruction like *

Akhaiu̯i̯oí

may perhaps not account properly for the Latin borrowing

Achīuī;

of course, the latter form may be simply the reflex of

Akhaioí

, with the

u

serving as hiatus breaker. Compare Latin Argīuī from

Argeîoi

(the Greek has no

u̯

before –

oi

); this Latin borrowing is attested early (e.g., Plautus), and I see no reason to insist on an analogical insertion of

u

by way of

Achīuī.

Another problem is that the reconstruction *

Akhaiu̯i̯ó

– would fail to account for *

Akhai̯u̯ó

-; this form, however, is not attested to my knowledge in Greek, unless we read the Cypriote spelling

ti-mo-wa-na-ko-to

/

sa-ka-i-wo-se

(Masson 1961n405.1) as *

Timou̯anaktosAkhaiu̯

os. This reading is vitiated, however, by the necessary assumption that word division has been neglected between the patronymic (genitive) and the hypothetical ethnic {353|354} (nominative). In fact, word divisions are faithfully observed in attested Cypriote spelling (word-final –

s

spelled –

se

). Also, there is an actual word divider between

ti-mo-wa-na-ko-to

and

sa-ka-i-wo-se

. Discussion in Masson 1961:69. Besides, etymologically genuine

u̯o

can be spelled

o

in Cypriote (Thumb/Scherer 1959:160), and we may therefore expect the reverse as well (

u̯o

for

o

).If indeed

Akhaió

– was never *

Akhaiu̯ó

-, then an argument could be made for its morphological parallelism with

krataió

– even without positing compound formations. {354|}