C. F. Burney wrote an article that has garnered attention in some circles: "Christ as the APXH of Creation," JTS 27 (1926) 160–177.
This article submitted that Colossians 1:15 exposits the first word of Genesis 1:1 (doing so midrashically), but accomplishes this job indirectly by alluding to Proverbs 8:22.
[See also Jeffrey S. Lamp, "Wisdom in Col 1:15–20: Contribution and Significance," JETS 41/1 (March 1998) 45–53.]
While the scholarly community at large has spoken approvingly of Burney's thesis, as with other matters, not all concur with his overarching thesis:
"This thesis [of Burney's] would presuppose that the passage was an exegesis of the Hebrew text, but the insight that 1:15-20 is a citation of a Hellenistic Christian hymn does away with this assumption. Moreover, it cannot be carried through in particulars without the aid of artificial explanations, and it is not sufficient for comprehension of the whole context--for this would be necessary if ARXH (beginning) from the second strophe is to be included. For a critique, cf. Jervell, Imago Dei, p. 200, n. 107; Gabathuler, Jesus Christus, Haupt pt der Kirche, 26-29; Feuillet, Le Christ sagesse, 189-91" (Eduard Lohse, Colossians and Philemon, page 47).
What Lohse means with respect to the "hymn" of Colossians 1:15-20 is that if we do possibly have what amounts to a pre-Pauline hymn in these verses, then it would seem unlikely that Paul is explaining a Hebrew text (as Burney suggests) since the hymn would evidently have Hellenistic as opposed to Semitic textual roots. Moreover, other NT "hymns" like Philippians 2:5-11 or 1 Timothy 3:16 would also have Hellenistic origins. That's provided these passages are truly hymnic.
But another criticism I have of Burney's work is that I believe he committed what Carson would later call "exegetical fallacies." One such fallacy is known as illegitimate totality transfer (as James Barr labels it). Neither RESHITH nor ARXH probably have all the meanings in any one context of usage (USUS LOQUENDI) that Burney attributes to each term. It's usually fallacious to believe that language works that way.
I do not think that the Midrashim always or necessarily engage in positing several different meanings for one word in a determinate context. The Midrashim may offer commentary on different levels of meaning that a word or verse has, but they do not have to posit multiple senses for words used in concreto. What I am saying is that Burney suggests RESHITH/ARXH possibly mean (i.e. lexically denote) "in RESHITH" "by RESHITH" "into RESHITH" "beginning" "Sum-total" "head" and "firstfruits" all in one context. This approach seems to be an example of illegitimate totality transfer. Midrash as an explanatory approach does not commit the exegete to the so-called ILL. See Burney, p. 176-177 and http://www.sacred-texts.com/jud/tmm/tmm07.htm
This post is urging that Burney is probably wrong in his analysis of RESHITH/ARXH because we now know that language generally does not operate in the way he contends. I do not even think that one who does Midrash is committed to the view that some Hebrew or Greek word in Scripture must bear multiple senses all at once within a given context.
I am not primarily trying to account for what the apostle Paul may or may not have done. My point is that it's usually fallacious to maintain that a word means S1, S2, S3, S4, S5, Sn in any given literary context where "S" represents a given "sense" for a particular sememe/morpheme/lexeme.