Phil 1:28 hHTIS (longish)

Iver Larsen iver_larsen at
Sun Jan 13 16:19:29 EST 2002

Dear Carl,

Since the issue is important beyond Phil 1:28 as you say, let me try to
bring us a bit further if I can. Thanks for your comprehensive response
which I have read with interest.
I may end up accepting your analysis, but I am still looking at an
alternative analysis which seems to me to be more consistent with the rest
of Greek grammar and less arbitrary. I'll make reference to the
Blass/Debrunner (BD) grammar which is the only one I have access to at the
moment, since most of my books are in another country where I normally live.
Your analysis is found in their paragraph 132 which states that if the
pronoun is subject, it agrees with the predicate noun, but not as
consistently as in Latin. I don't know how inconsistent this is in Greek.
When an analysis suggests inconsistencies there is at least the chance that
an alternative analysis is also possible. The alternative analysis would be
based on BD 296, see below.

> I'm going to make one more response on this matter at
> this point, since (a) I think we've reached an impasse and remain in
> serious disagreement over important matters, and (b) I really need to
> consult some hard-copy reference works to which I won't have
> access until I return to my NC home at the end of this month before I can
respond more
> adequately to some of the challenges.
> What I would
> affirm is that (a) the antecedent of a relative pronoun/clause is to be
> sought in the text which precedes it and its clause, (b) that generally
> this antecedent is a noun with which the relative pronoun clearly "agrees"
> in gender and number, but (c) the relative pronoun may occasionally have a
> demonstrative force and, although referring to content that
> precedes it and its clause, may take its number and gender from a
"postcedent" substantive
> that may be understood as either its subject or predicate word. In such an
> instance, I think, we might say that the relative pronoun is
> functioning as a demonstrative, and that its sequential and logical sense
in its context
> is, "and this/that is {x}. And I think that when this happens the
> principle of concord is pretty much that for a demonstrative in a similar
> e.g.:
> 	Jn 3:19 hAUTH DE ESTIN hH KRISIS ...
> Sometimes the demonstrative may be anaphoric to what has been stated in
> preceding discourse, and still take its concord from the postcedent noun.
> For instance, in Mt 7:12,
> hOUTOS  refers back to the moral principle that we now term "the golden
> rule," but it takes its gender and number from the first of the nouns
> following in the group hO NOMOS KAI hOI PROFHTAI.

I agree with your (a) and (b), but would add something to (b), namely that
at times the relative does not agree with a specific noun in the preceding
context but with an implied noun which can be associated with that context
(paragraph 296 in BD talks about this constructio ad sensum for the relative
You have managed to find one instance of hOUTOS out of many others that does
seem to support your thesis (and BD 132). But I think an alternative
analysis is possible, the constructio ad sensum that I just mentioned. The
ESTIN is not here a marker of equivalence and to translate it with English
"is" is not very meaningful. NIV says "for this sums up the Law and the
Prophets". TEV says "this is the meaning of the Law of Moses and of the
teachings of the prophets." The word "this" seems to refer back to the
golden rule, as you say, or this principle (NOMOS) which may be derived in
some way from the OT, that is, the Law (also NOMOS) and the Prophets.

I could give some other examples like:


I agree that the ordinary and indefinite relative pronoun can be equivalent
to a demonstrative+relative, but I think one needs to be careful
transferring what applies to the pure demonstrative to the dem+rel. The
dem+rel I would take as a subclass of the rel, not a subclass of the dem. I
would like to see examples of the ordinary relative pronoun that also take
their gender and number from a postcedent noun. That would help to persuade

Clearly hOUTOS is occasionally kataphoric, referring to a noun that follows
it, and in such cases it agrees with that noun. If I understand you
correctly, your point (a) says that the rel is always anaphoric, and if so,
it is in this respect not comparable to the demonstrative.
> I do want to add here by way of a footnote, that hO/ TI does in
> fact appear in the GNT after all, as Steven Lo Vullo has pointed out to me
> treats it as two words rather than as a form of hOSTIS; I don't know how
> other programs may handle it). However, the information about this is not
> helpful; there are only 9 instances, (Mark 6.23; Luke 10.35; John 2.5;
> 8.25; 14.13; 15.16; Acts 9.6; 1Cor 16.2; Col 3.17). Most of these
> appear in generalizing constructions with EAN + subjunctive, and without
> every instance is to be understood as indefinite relative pronoun with
> implicit embedded antecedent in the sense of "that which" or "whatever."

Yes, in my program these were treated separately as hO and TI where the hO
was then to be found under the hOS key entry rather than hOSTIS. But this is
quite different in several respects. One of the obvious differences is that
there doesn't seem to be any clear hOSTIS forms in the GNT outside of the
nominative (barring the fixed form hEWS hOTOU), and all of the hO/ TIs above
are in the accusative, so hO/ TI does not occur in the nominative in a
construction comparable to what we are looking at.
> >For 1 Cor 3:17 which has hOITINES I can substitute with "these people" or
> >"such people", referring to the people who constitute NAOS QEOU.
> This doesn't really make sense to me, Iver; the text is hO GAR NAOS TOU
> QEOU hAGOIS ESTIN, hOITINES ESTE hUMEIS. I don't see how "temple" can
> become "these people" without the explicit clarification of the relative
> clause, which I would English as "which (temple) is what you are"--but the
> form hOITINES clearly is determined by the form of the
> pronoun hUMEIS.

Here we disagree as to what makes sense, because we differ in our
presuppositions. But let me quote from BD 296 on constructio ad sensum:
They give Lk 6:17 as an example where the relative hOI refers back to the
singular PLHQOS. The sense of PLHQOS is "many people". They give Phil 3:20
as en example of a relative singular EX hOU referring back to OURANOIS,
which although in plural formally, can easily be considered a singular
concept. Then they give Gal 4:19 TEKNA MOU, hOUS where hOUS refers to
"people" although the specific word is TEKNA.
> >For Eph 3:13 which has hHTIS, I can substitute "this affliction"
> (The idea probably is that you should feel honoured that we are willing to
> suffer for you).
> This remains disputed, and I'll comment on it at the point where you bring
> this passage up again later.

Again, I would take this with BD 296.
> >For Phil 1:28 which has hHTIS, I can substitute "this faith" and
> make very good sense of it - see below.

This I would not take with BD 296, because I cannot find any concept in the
immediately preceding context to associate with a feminine form. So, I take
it back to a specific feminine word.

> What I fail to see is the probative value of these cited versions. I'd say
> that even if NRSV uses "they" and understands the hHTIS as referring back
> to QLIYEIS, that doesn't resolve the question whether hHTIS takes its
> gender from an implicit singular QLIYIS, as you assert, or from its
> predicate noun DOXA, as I prefer to think. And so with the other versions:
> they are translating the SENSE as each committee understands it, not the
> grammatical construction.

If they translate the SENSE, could the Greek then have a constructio ad

Thanks for listening to my probing of controversial grammatical ideas,

Iver Larsen

More information about the B-Greek mailing list