A Response to David K. Bernard on John 1:1

Two gentlemen who call themselves “The One Dollar Apologists” recently interviewed David K. Bernard about the Oneness Pentecostal view of church history and of the Scriptures. Dr. Bernard currently serves as the General Superintendent of the United Pentecostal Church International (UPCI), and is a monumental figure in the UPCI for several reasons.

In part two of their interview, they discuss a number of scriptures and how they relate to the Oneness view of Jesus Christ. In this post I’m only going to focus on what Bernard says about John 1:1. This is what that verse says in both Greek and English:

In the beginning was the Word, and the Word was with God, and the Word was God.

Ἐν ἀρχῇ ἦν ὁ λόγος, καὶ ὁ λόγος ἦν πρὸς τὸν θεόν, καὶ θεὸς ἦν ὁ λόγος. (NA28)

This verse makes three statements, each of which is separated by a comma in the English above. Scholars sometimes refer to these statements as John 1:1a, John 1:1b, and John 1:1c, respectively. That’s how I’m going to refer to each of these parts of the verse in this post.

Bernard makes two points that I’d like to address:

    1. A Trinitarian interpretation of John 1:1 requires that we change the definition of the word “God” in the middle of the verse.
    2. The preposition “with” (πρός) in John 1:1b can mean “pertains to,” and counts in favor of a Oneness understanding of the verse.

As I hope to show, the first claim is inconsequential on a proper understanding of John 1:1, and the second claim is grammatically mistaken.

“God” and “Was God”

With regard to the first point, here’s what Bernard says (my transcription):

If you take a Trinitarian interpretation, the only way it fits is [if] you have to change the definition of “God” in the middle of the verse. Here’s what I mean: If you say, “Well, God is a Trinity,” are you willing to say, “In the beginning was the Word, [and] the Word was with the Trinity, and the Word was the Trinity”? Well obviously that would demolish Trinitarianism.

So, the Trinitarian would say, “No, normally ‘God’ is ‘God the Father'” (as I alluded to earlier). So, “In the beginning was the Word, [and] the Word was with God”—so [the Trinitarian would say,] “See, the Word (the second Person) was with God (the first Person).” But then, “the Word was God.” So are you willing to translate [“God”] consistently throughout the verse? “In the beginning was the Word, [and] the Word was with God the Father, [and] the Word was God the Father”? Well, then you’re back to pretty much my explanation.

So, the only way Trinitarianism works is [when] you have to have this “P.S.” assumption: “In the beginning was the Word (second Person), [and] the Word was with God (the first Person), and the Word was God (the second Person).” So you’re actually changing the definition of “God” in the middle of the verse to make it fit the doctrine of the Trinity. Is that really what John had in mind?

For those who don’t know Greek, what Bernard is asserting here can easily pass by without notice. The thing is, Bernard assumes, and doesn’t argue for the claim, that John 1:1c is an identity claim rather than a predicative claim.

To begin to see where Bernard goes wrong, consider this reconstruction of his argument:

  1. Suppose that God is numerically identical to the Trinity (or God = Trinity).
  2. Therefore, the Word = the second Person of the Trinity.
  3. Either “God” means “the Trinity,” “the Father,” or “the Son.”
  4. If “God” means “the Trinity,” then the Word was numerically identical to the Trinity (i.e., Word = the Trinity).
  5. The the Word = the Trinity, then it is not the case that the Word is the second Person of the Trinity (i.e., John 1:1c “demolishes Trinitarianism”).
  6. Therefore, if “God” means “the Trinity,” then it is not the case that the Word is the second Person of the Trinity.
  7. If “God” means “the Father,” then the Word was numerically identical to the Father.
  8. If the Word was numerically identical to the Father, then is not the case that the Word is the second Person of the Trinity. (Suppressed premise)
  9. Therefore, if “God” means “the Father,” then is not the case that the Word is the second Person of the Trinity.
  10. If “God” means “the Son,” then “God” in John 1:1c has a different meaning from “God” in John 1:1b.
  11. If “God” in John 1:1c has a different meaning from “God” in John 1:1b, then the Trinitarian understanding of John 1:1 is ad hoc.
  12. Therefore, if “God” means “the Son,” then the Trinitarian understanding of John 1:1 is ad hoc.
  13. Therefore, either the Word is not the second Person of the Trinity, or the Trinitarian understanding of John 1:1 is ad hoc.

Why think each of the steps in this argument are true? Premise (1) is assumed for the sake of argument and needs no justification, and the conclusion at step (2) follows from what Trinitarians affirm about God. For now let’s grant the truth of premise (3), but note that this premise (and those following) have to do with the meaning of “God” in John 1:1c.

The conclusions at steps (6), (9), and (12) follow from a rule of deductive logic called “hypothetical syllogism.” The conclusion at (13) follows from (3), (6), (9), and (12) from another rule called “constructive dilemma.” Obviously, what Bernard is trying to do at (13) is show that the Trinitarian interpretation of John 1:1 is problematic no matter how one understands “God” there. It either leads to a contradiction, or else says that “God” means “God the Son” and so essentially reads into the text a solution that stems from Trinitarian theology. It is “made up” to save from contradiction, which is what ad hoc means here.

Now, Bernard thinks that premises (4), (7), and (10) are true because, as he says in the interview (my transcription), “And actually the [word] order—the Greek [word] order—is emphatic. You can actually translate it, as The Amplified [does], as ‘the Word was God himself.'” In other words, for Bernard, John 1:1c means that the Word is numerically identical to whatever “God” refers to.

It seems to me that Bernard should know better. Syntactically, John 1:1c contains a construction called a “Colwell Construction.” (Syntax refers to how words are organized.) But what’s that? Take a closer look at the word “God” (θεὸς) in John 1:1c:

 [a] Ἐν ἀρχῇ ἦν ὁ λόγος, [b] καὶ ὁ λόγος ἦν πρὸς τὸν θεόν, [c] καὶ θεὸς ἦν ὁ λόγος.

There are three things true the word “God” there.

  1. It occurs before the Word “was” (ἦν), and is therefore pre-verbal.
  2. It lacks the definite article (“the”; Gr. ), which is what “anarthrous” means.
  3. It is in the nominative case and functions as a predicate with relation to “the Word” (ὁ λόγος). This is what’s called a “predicate nominative.” 

So, a “Colwell Construction” is a case where we have a pre-verbal, anarthrous predicate nominative.

The syntax of a Colwell Construction has two semantic possibilities, or meanings. The first is what I referred to above as an identity claim, and what Daniel B. Wallace calls a “convertible proposition” in Greek Grammar Beyond the Basics. This is a case where the anarthrous predicate nominative is numerically identical to the subject of the Colwell Construction. Here’s an example:

Nathaneal answered him, “Rabbi, you are the Son of God, you are the king of Israel.” (John 1:49)

ἀπεκρίθη αὐτῷ Ναθαναήλ· ῥαββί, σὺ εἶ ὁ υἱὸς τοῦ θεοῦ, σὺ βασιλεὺς εἶ τοῦ Ἰσραήλ

On this verse Wallace says:

Nathanael’s response to Jesus is a twofold identification. In the first construction the PN [i.e., predicate nominative] follows the verb and has the article. In the second construction the PN precedes the verb and lacks the article. (p. 263)

In the second of the two identifications, Nathaniel also says that Jesus is the king of Israel. But, as Wallace points out, “king” (βασιλεὺς) is anarthrous in Greek. So why do translators add the definite article in translation? It’s because, in context, Nathanial doesn’t seem to be saying Jesus is king of Israel, nor is he saying that Jesus has the qualities of a king. Nathaniel seems to be saying Jesus is the king of Israel (of which there’s only one anyway).

If Nathaniel were asserting that Jesus were king, the nominative “king” (βασιλεὺς) would bean indefinite predicate. If he were saying Jesus had the qualities of a king, “king” would be a qualitative predicate. But to say Jesus is the king makes “king” a definite predicate. A definite noun is a noun that either has the definite article (or is articular), or functions as if it were articular.

This helps us see why Nathaniel is making an identity claim. The proposition he is affirming here is “Jesus is the king of Israel,” where “king” is a definite predicate. Jesus is numerically identical to the king of Israel, which means Jesus = the king of Israel. But identity is a symmetrical relation. So it is equally true to say that Nathaniel is affirming that “the king of Israel = Jesus.” If we swap the positions of “the king of Israel” and “Jesus” in this identity claim, Nathaniel is still making the same claim. So “Jesus” (the referent of σὺ) and “the king of Israel” (βασιλεὺς . . . τοῦ Ἰσραήλ) are convertible terms, and we have a convertible proposition. Nathaniel is effectively saying σὺ εἶ ὁ βασιλεὺς.

The second semantic possibility for a Colwell Construction is a predicative claim, or what Wallace calls a “subset proposition.” Here’s an example everyone knows:

God is love. (1 John 4:8)

ὁ θεὸς ἀγάπη ἐστίν

Wallace comments:

The meaning [of 1 John 4:8] is certainly not convertible: “love is God.” The idea of a qualitative ἀγάπη is that God’s essence or nature is love, or that he has the quality of love. Thus love is an attribute, not an identification, of God. (p. 264)

In other words, John is not saying that God is love itself, where “love” refers to some abstract universal. Rather God is in a subset of beings that are loving beings. (Obviously, on the Christian view, God is the most loving being of all.)

Okay, now we’re in a position to see how this relates to John 1:1c, using the technical vocabulary we’ve learned so far. To state what I said above differently now, Bernard assumes that “God” (θεός) in John 1:1c is a definite predicate nominative, and that what’s going on there is a convertible proposition. The problem is that it is far more likely that “God” (θεός) is a qualitative predicate in John 1:1c.

Dr. Donald Hartley has put in the legwork to support this claim. What he says about this is in-depth and technical, so I’ll only summarize his key points so that anybody can understand them.

Hartley provides statistics for all of the Colwell Constructions in the New Testament. The relevant chart for our purposes provides an answer this question: “If all definitizing factors [e.g., genitive adjuncts, monadic nouns, or proper names] are omitted from singular count nouns, what semantic predominates?” Really, the question asks this: When considering any particular Colwell Construction in the New Testament that is relevantly similar to John 1:1c, what meaning does the predicate nominative carry? Hartley answers that only 3% (2 of 66 occurrences) of these have definite predicate nominatives. In his study, he excluded disputed passages (like John 1:1c!) from these statistics.

What’s more relevant, though, is how the Gospel of John uses Colwell Constructions. Hartley has a figure for this as well. He concludes that in only 11% (or 2 of 18) of the constructions relevantly like John 1:1c, the predicate nominatives are definite. Yet in 56% percent (or 10 of 18) of them the predicate nominatives are qualitative. Again, he excluded John 1:1c from these statistics.

I find John’s usage of Colwell Constructions decisive on its own. But no matter which way we take the statistics, I don’t see how anyone can reasonably conclude that θεός in John 1:1c is definite. It is indeed an “improbable venture,” as Hartley says, to make the assumption that Bernard does. “God” (θεός) in John 1:1c is least likely definite, and therefore John 1:1c is least likely a convertible proposition. John probably isn’t making an identity claim in John 1:1c, as Bernard would have us believe. Rather, he’s making a predicative claim by using θεός as a qualitative predicate, and we’re actually dealing with a subset proposition.

All of this means that we can deny premise (3) of the reconstruction of Bernard’s argument. The justification for premise (3) is something like the following sub-argument:

3.1. If John 1:1c is a convertible proposition, then either “God” means “the Trinity,” “the Father,” or “the Son.”
3.2. John 1:1c is a convertible proposition.
3. Therefore, either “God” means “the Trinity,” “the Father,” or “the Son.”

I’ll grant (3.1). But as we’ve seen from this section, there’s little reason to think premise (3.2) is the case; it is more probably false than true. We can therefore avoid Bernard’s argument.

I think Bernard is right that, if “God” is definite in both John 1:1b and 1:1c, there’s a problem for Trinitarianism. In order to avoid the problem, there would have to be an equivocation on “God” to save Trinitarianism: “and word was with God (the Father), and the Word was God (the Son).”  But if “God” is a qualitative predicate in John 1:1c, the fact that the meaning of “God” changes is inconsequential. This is, after all, John’s use of the term in this verse, as it goes from being a definite noun in John 1:1b to a qualitative one in 1:1c. If John wants to use “God” differently from one part of the verse to the other, then we should let John be John.

“With God”

This is what Bernard says about the phrase “with God” (πρὸς τὸν θεόν) in John 1:1b (my transcription):

If we could go back to the first century—to the Hebrew and Greek context that I explained earlier—I think we would understand [that,] “In the beginning was the Word, [and] the Word pertained to God, and the Word actually was God.”

Notice that Bernard takes “with” (πρός) to mean “pertain to.” He points out in the interview, as he does in The Oneness of God, that Hebrews 2:17 uses πρός followed by the accusative case and (according to the KJV translators) means “pertained to.” He says:

Note also that the Greek word pros [πρός], translated “with” in [John 1] verse 1, is the same word translated “pertaining to” in Hebrews 2:17 and 5:1. John 1:1 could include in its meanings, therefore, the following: “The Word pertained to God and the Word was God,” or “The Word belonged to God and was God.”
—David K. Bernard, The Oneness of God, p. 61

All of this is mistaken. The grammar in Hebrews 2:17, 5:1, and Romans 15:17 (also translated this way) doesn’t match what we have in John 1:1b. Namely, in each of those verses, there is a definite article that precedes πρός. Here are those verses in English and Greek:

Therefore he had to be made like his brothers and sisters in every respect, so that he could become a merciful and faithful high priest in things relating to [KJV: “pertaining to”] God, to make atonement for the sins of the people. (Hebrews 2:17, NET)

ὅθεν ὤφειλεν κατὰ πάντα τοῖς ἀδελφοῖς ὁμοιωθῆναι, ἵνα ἐλεήμων γένηται καὶ πιστὸς ἀρχιερεὺς τὰ πρὸς τὸν θεὸν εἰς τὸ ἱλάσκεσθαι τὰς ἁμαρτίας τοῦ λαοῦ. (NA28)

For every high priest is taken from among the people and appointed to represent them before God [KJV: “in things pertaining to God”], to offer both gifts and sacrifices for sins. (Hebrews 5:1, NET)

Πᾶς γὰρ ἀρχιερεὺς ἐξ ἀνθρώπων λαμβανόμενος ὑπὲρ ἀνθρώπων καθίσταται τὰ πρὸς τὸν θεόν, ἵνα προσφέρῃ δῶρά τε καὶ θυσίας ὑπὲρ ἁμαρτιῶν (NA28)

So I boast in Christ Jesus about the things that pertain to God. (Romans 15:17, NET)

ἔχω οὖν [τὴν] καύχησιν ἐν Χριστῷ Ἰησοῦ τὰ πρὸς τὸν θεόν· (NA28)

Clearly, πρός has the meaning in these verses that it does because the prepositional phrase is articular. What reason is there, then, to take John 1:1b the way that Bernard does? Other than his mistaken examples, Bernard simply doesn’t provide any. We’re going to need better reasons, and better examples, to think that πρός really means “pertains to” in John 1:1b.


Is Bernard’s understand of John 1:1 a possible interpretation? Of course it is, and I haven’t once said that it impossible. What I have argued, however, is that the reasons Bernard gives in favor of his understanding of the verse are either inconsequential (as in John 1:1c), or grammatically mistaken (John 1:1b). I have offered undercutting defeaters to the claims that Bernard makes in his interview.

There’s a lot more to discuss about John’s prologue, and John 1:1 in particular. I’ll have a lot more to say over time.

2 thoughts on “A Response to David K. Bernard on John 1:1”

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s