Dear Nina

You wrote:

> Dear Yong Peng,
> As Gunnar said, pi: also, I think.
> But I would say: I go to the Buddha for refuge. Or: I take my refuge
> in the Buddha.
> We have two accusatives, and the Co (to the Khuddakapaa.tha) I quoted
> talks at length about these. They say that gacchami does not take two
> accusatives. I also find it difficult to apply here Relational Grammar.

It is in such a case that we have to rely on commentaries. I have chosen
a slightly different context here because the commentary is much clearer
than that of Khuddakapaa.tha and definitely a genuine work of
Buddhaghosa. I won't claim that the explanation of Buddhaghosa is
absolutely correct but, at least, we should know how he understood such
a usage.

esaaha.m bhavanta.m gotama.m sara.na.m gacchaami dhamma~nca
bhikkhusa'ngha~nca. (Vin-3.6)

We would go to the commentary first.

1. esaahanti eso aha.m. (Sp - 1.171) (= The phrase "esaaha.m" means
"That I")
2. bhavanta.m gotama.m sara.na.m gacchaamiiti bhavanta.m gotama.m
sara.nanti gacchaami. (ibid) (= The sentence "bhavanta.m gotama.m
sara.na.m gacchaami" means " (I) rely on the Venerable Gotama as refuge)

(1) is not so difficult to translate but we would miss the commentator's
real intent if we don't notice that he indicates by (1) the fact that
"essaha.m" is a Sandhi combination of "eso" and "aha.m".

>From (2) we know that "iti" is understood after "sara.na.m". In other

words, "sara.na.m" is an idiomatic representation of "sara.nanti". The
word "sara.nanti" can be viewed as an indeclinable compound having the
instrumental case and related to "gacchaami" in Adverbial (ADV)
relation. But the word "sara.na.m" itself is of nominative case.

Then the sentence can be translated as "That I rely on the Venerable
Gotama as refuge".

with metta

Ven. Pandita