A DECISIVE FACTOR FOR THE VALIDITY OF THE FORMULA
PROTOTOKOS + PASIES KITISEOS = PARTITIVE GENITIVE
Luis C. Reyes originally penned this challenge in a summarized form to post on the Greek Theology discussion board, moderated by Edgar Foster, one of Jehovah’s Witnesses. Mr. Foster did not approve the original message, stating that he felt this topic had been thoroughly discussed already on the board. Mr. Reyes felt that though he had discussed this topic on other boards, it had not been addressed on Greek Theology. In any event, with Mr. Reyes’ permission, I am posting his challenge here. If anyone would like to respond to Mr. Reyes, he or she may contact For an Answer here. The original message has been expanded and incorporated with additional information that was not included in the original challenge.
In this short article I challenge the notion that the ancient koine Greek term prototokos intrinsically possessed a partitive semantic value encoded in its linguistic form. Some advocates of the “partitive word” view, for instance Wes Williams, have proposed that the proper interpretation of prototokos pasies kitiseos ( Col 1:15 ) can be obtained by pure linguistic decoding of the formula: prototokos (partitive word) + pasies kitiseos (genitive construction) = partitive genitive construction (i.e., the prototokos is included in the class of creation). Theological conclusion: The Son is a part of creation. While I do not agree with the theological conclusion drawn from such a linguistic formula, my point of contention here is purely a linguistic one. In fact, the entire argument hinges on the critical issue of whether or not the isolated linguistic component prototokos possesses an intrinsic partitive semantic value. From the perspective of modern linguistics and cognitive pragmatics, I argue that it is perhaps impossible to prove that an isolated linguistic component (such as prototokos) can convey an intrinsic partitive semantic value. As a result, I argue that the entire formula proposed by Williams be abandoned, since there has been no linguistic evidence to substantiate the “partitive word” view with prototokos, which is critical for the application of the formula as a whole.
There have been many who have spent a great deal of time attempting to obtain the proper interpretation of Col 1:15, specifically the phrase prototokos pasies kitiseos. Some who have understood prototokos to be a part of creation have understood the genitive of the whole construction to be a partitive genitive. Debate has arisen with encounters with people of the Jehovah’s Witness persuasion on various Internet forums over this issue. Here I will only focus on refuting one particular aspect of the argument that has apparently received much prestige and assurance among some Jehovah’s Witnesses. This argument has come to be known to me as the “partitive word” argument, and—to the best of my knowledge—Wes Williams first proposed it after refining a similar argument originally proposed by Rolf Furuli.[i]
- WES WILLIAMS, ROLF FURULI AND THE PROTOTOKOS/PARTITIVE WORD ARGUMENT
Williams has attempted to appeal to the notion that prototokos is intrinsically a “partitive word” in order to understand the proper interpretation of the grammatical construction prototokos pasies kitiseos in Col.1: 15. According to Williams in over 70 passages in the LXX every time the word appears in the genitive (excluding metaphorical uses) it always includes the prototokos in a particular class. Wes has stated:
I posted on this board about two years ago every non-metaphorical instance in the LXX where it occurs with a genitive (or “of a “) phrase. What the examples show is that the PRWTOTOKOS (“firstborn”) is a PART of the group.[ii]
With the data obtained from his analysis of prototokos and its uses in the LXX, Williams has concluded that prototokos and its genitive pasies kitiseos in Col 1:15 must necessarily be a partitive genitive. Williams also concludes from his analysis that prototokos must necessarily possess an intrinsic partitive semantic property. In fact, in August of 1998, in a post arguing against Dr. Robert Keay,[iii] Williams stated his proposed argument as:
I have been through all the occurrences of PRWTOTOKOS in the LXX. I present the following usage for proof behind my point: 27 examples of partitive genitive (the firstborn is a part of the group): Gen 4:4; 25:13; Ex 11:5; 13:13,15; 22:28; 34:19,19; 34:20, 20; Num 3:40, 41, 41;3:45, 46, 50; 8:16; 18:15, 15; Deut 12:6, 17; 14:23; 15:19; Neh 10:37, 37; Ezek 44:30. 42 examples of possessive genitive, such as `my son`,implying membership of the group of sons: Gen 49:3; Ex 4:22; 4:23; 6:14;11:5; Num 1:20; 18:17,17,17;26:5; Deut 21:15,16,17; 33:17;Judg 8:20;2 Sam 3:2; 2 Sam 13:21; 1 Kings 16:34; 1 Chr 1:29; 2:3,13; 2:25,25,27,42,50; 3:1,15; 4:4; 5:1,3; 8:1,30,38,39; 9:5,31,36,44; 26:2; Psalm 134:8; Mica 6:7; Jer 38:9 There are no example [sic] of other genitives. Lexical semantics, therefore, sans theology, give one meaning to PRWTOTOKOS, and this meaning is intrinsically partitive. Philologically speaking, all genitives with the word uphold the partitive meaning.[iv]
Now before discussing any of the content of this citation there are a few other things that need to be mentioned. First, it appears that William’s argument noted above was possibly taken from another discussion forum where Rolf Furuli actually originated the argument. In fact, in March of 1997 (almost half a year prior to Williams citation above), Furuli posted:
I have been through all the occurrences of PRWTOTOKOS in the LXX with the following results. 27 examples of partitive genitive: Gen 4:4; 25:13;Ex 11:5; 13:13,15; 22:28; 34:19,19; 34:20, 20; Num 3:40, 41, 41; 3:45, 46, 50; 8:16; 18:15, 15; Deut 12:6,17; 14:23; 15:19; Neh 10:37, 37; Ezek 44:30. 42 examples of possessive genitive, such as `my son`, implying membership of the group of sons: Gen 49:3; Ex 4:22; 4:23; 6:14;11:5; Num 1:20; 18:17, 17,17; 26:5; Deut 21:15, 16, 17; 33:17; Judg 8:20; 2 Sam 3:2; 2 Sam 13:21; 1 Kings 16:34; 1 Chr1:29; 2:3, 13; 2:25, 25, 27, 42.50; 3:1, 15; 4:4: 5:1, 3; 8:1 ,30 ,38 ,39; 9:5, 31, 36, 44; 26:2; Psalm 134:8; Mica 6:7; Jer 38:9. There are no example [sic] of other genitives. Stage I: Lexical semantics, therefore, sans theology, give one meaning to PRWTOTOKOS, and this meaning is intrinsic partitiv [sic]. Philologically speaking, all genitives with the word uphold the partitive meaning.[v]
The point I wish to make here is not that William’s comments are similar to Furuli’s, but rather that this “partitive word” argument with prototokos (as it is described above) was possibly first originated by Furuli, since it predates William’s comments.[vi] Now, while Furuli does not explicitly say that prototokos is a so-called “partitive word,” he does state however (as does Williams), that lexical semantics give ONE meaning to prototokos, which they both say is “intrinsically partitive.”[vii]
Also, elsewhere Furuli notes that although he does not object to the use of the phrase, “a partitive word” in a descriptive way, he does nonetheless argue that the very lexical meaning of prototokos makes the term a part of the whole. This means for Furuli that prototokos intrinsically conveys a partitive semantic value:
While I am not aware of a formal linguists use of the term “a partitive word,” I do not object to using it in a descriptive way. I do not say it is wrong to use this expression, I just say it is not a standard linguistic expression. The lexical meaning of certain words imply a certain relationship to other words, such as “shepherd” and “flock”, “king” and “queen”, “salt” and “pepper” etc. While such relationships are not “semantic”, they give strong indications of the meaning of grammatical relationships that generally are ambiguous, such as Greek genitive […] The English first-born means “one who is born first”, and the Greek word PRWTOTOKOS is used throughout the LXX and the NT with the same meaning; there is no example of a different lexical meaning. So the use of PRWTOTOKOS implies the existence or possible existence of a group of creatures of the same kind as the one who is PRWTOTOKOS. When we have a genitive relationship between PRWTOTOKOS and such a “group” word, the genitive is partitive. The very lexical meaning makes the word a part of the whole in such a situation.[viii]
This is a notion that is often clearly repeated among advocates of this view. For instance, Williams has also explicitly stated, “The word PRWTOTOKOS (`firstborn`) is a partitive word. It has an intrinsic partitive force.”[ix] In another place Williams has further stated:
The key to bear in mind is that the partitive word PRWTOTOKOS (firstborn) lexically requires that the firstborn be a part of the group. This is a direct statement that Christ is part of the “creation” in Col 1:15. This is clear, powerful, and fully lexically supported by the LXX![x]
Therefore, the first premise of this argument is that prototokos possesses an intrinsic partitive semantic value inherent in the isolated word. According to Williams this partitive force “lexically requires that the firstborn be a part of the group.” However, the argument is expanded when this factor is combined with a genitive construction to produce a working linguistic formula for Williams.
- THE FORMULA FOR PROPERLY UNDERSTANDING PROTOTOKOS PASIES KITISEOS.
Williams writes, “Theology aside, I think the grammar of PRWTOTOKOS + genitive is inescapable for a partitive genitive or genitive of relation.”[xi] Hence, Williams has proposed the below listed formula for the proper interpretation of prototokos pasies kitiseos (Col 1:15):
|Prototokos/ partitive word||+||Paseis kitiseos/ genitive construction||=||Partitive genitive construction as a whole|
A critical linguistic variable is overlooked in the argument: To the best of my knowledge, no one has ever presented a shred of linguistic evidence to support the notion that the isolated linguistic component prototokos is intrinsically a so-called partitive word (possessing an intrinsic partitve semantic value on its own), and certainly, no linguistic evidence has been demonstrated for the notion that prototokos “lexically requires that the firstborn be a part of a group.” It would seem that such an interpretation might be derived from the extra-linguistic context or from pragmatic implicature (although I even dispute that), but where is the linguistic evidence to substantiate such a claim of an intrinsic semantic partitive value inherent in the linguistic component prototokos?
This is very important because if the proper formula for interpreting Col 1:15 is: prototokos (partitive word) + pasies kitiseos (genitive construction) = partitive genitive construction, then demonstrating that prototokos is a so-called, “partitive word” is a critical variable for the validity and application of the formula to Col 1:15 (then one would have to consider the entire argument as a whole). In fact, in order to even begin applying this formula to Col 1:15 , advocates of this view first need to provide linguistic evidence that the isolated prototokos is a so called “partitve word” to begin with. The proposed linguistic formula is meaningless without linguistic justification for this critical element in the formula. If advocates of this view first cannot establish linguistically that prototokos is a so-called “partitive word,” then the entire argument collapses because it does not meet its own specifications and criteria for the proposed linguistic formula for interpretation.
- A CHALLENGE TO PRODUCE LINGUISTIC EVIDENCE FOR THE CLAIM
(a). It should be noted that my objection is strictly a linguistic one. Since I question the notion for the existence of an intrinsic partitive semantic value in an isolated linguistic component (it is an assumption that first needs to be proven), then advocates of the “partitive word” view first need to provide some kind of statement of the methodological procedures that they have utilized which have lead them to identify an intrinsic partitive semantic value in the isolated linguistic component, prototokos. I am not asking for evidence from the micro pragmatic or extra linguistic context. Neither am I asking for linguistic evidence from the grammatical-syntactical context involving a string of other linguistic items. Since it is argued that prototokos conveys this intrinsic partitive semantic value, it is specifically this term for which I am requesting evidence. Hence, if there is any such linguistic evidence it must be demonstrated from the lexical semantics of the isolated term.
(b). It must clearly be articulated if this so-called partitive force is determined to be intrinsic in the isolated linguistic component itself, or if it is transported and conveyed over to the linguistic item prototokos as an implicature from either the pragmatic, extra linguistic context, or from any other outside influence. Notice, if it is argued that the partitive force is conveyed over to the linguistic term as an implicature, this does absolutely nothing to prove the notion that prototokos conveys an intrinsic parititve semantic value standing on its own. If this were the case the entire proposed linguistic formula collapses.
(c). If it is argued that the partitive force is determined to be intrinsic in the isolated encoded linguistic form of prototokos, advocates of this view are required to demonstrate from the isolated encoded linguistic form such intrinsic semantic value or property. How was this partitive semantic value determined and measured? What methodology was used to determine this intrinsic partitive semantic value so that others can replicate the procedure? If the procedures were legitimate and scientific (of course I am presupposing the scientific study of language here), then perhaps others can also reach similar results by utilizing the same methodology. One would expect that there would be some scientific procedure used to substantiate the view that prototokos possesses an intrinsic partitive semantic property. That is of course if a scientific approach to answer the question can even be applicable to this issue to begin with. Incidentally, Furuli has attempted to utilize scientific principles for linguistic analysis of the biblical texts. The back cover of Furuli’s book reads:
In the natural sciences, a basic principle is to break everything down to the smallest possible units and then study each unit. In linguistics and in the study of the biblical languages, a similar principle was followed with the word as the basic unit, but from the middle of this century the view has developed that the smallest units which were meaningful for translation had to be the sentence or even the paragraph. The author believes that the pendulum has swung too far in one direction, and that it still is meaningful to work with the word as the fundamental unit of translation.[xii]
I am of the opinion that a scientific method and approach, as the one Furuli’s work advocates, cannot be adequate to test the validity of the partitive word argument. While Furuli’s main concern above deals with translation, the principle that it is meaningful to work with the word provides no scientific support for the notion that prototokos conveys an intrinsic partitive property. The reason is because science focuses on either confirming or disproving a testable body of knowledge. But how does one test and confirm or disprove whether prototokos intrinsically conveys a so-called partitive semantic value inherent in its linguistic form? Science can certainly be described as a critical set of research methods designed to describe and interpret empirically perceived phenomenon. In fact, the scientific method possesses certain defined principles, such as the employment of methods of systematic empiricism. As noted, it also holds the concept that in order for something to be tested by science, it first must be a problem that is empirically solvable, and that it is capable of yielding testable theories. It is critical for the scientific method that its aim center solely for knowledge that is publicly verifiable and that it be open to verification or rejection by the process of replication. However, it seems to me that the notion that prototokos conveys an intrinsic partitive value is categorized as something that is non-testable, and hence out of the reach of scientific investigation. The notion that prototokos conveys an intrinsic partitive semantic value appears to be a problem that is not falsifiable, and thus is unlikely to be altered or shown to be false. Therefore, the burden is on those who advocate this position. This is precisely the reason why I have requested advocates of the partitive word view to prove their case from the isolated term. From a purely linguistic perspective, no linguistic evidence has been provided from the lexis itself to substantiate the notion that prototokos intrinsically conveys a partitve semantic value of which one can decode something like, “firstborn of a class.” It is critical to note that the “of a class” portion is not information linguistically encoded in the linguistic component, but rather if it is conveyed in a particular context, it may be understood from extra-linguistic or pragmatic implicature. Such an implicature may certainly be understood from a purely pragmatic perspective. However, this does nothing to prove the assumption that it was derived from the intrinsic isolated linguistic component itself. This is something that needs to be substantiated linguistically, and indeed, this is the basis of my entire criticism of the partitive word argument. I believe that the partitive word notion cannot be substantiated linguistically without crossing over into pragmatics and extra linguistic interpretation. Consequently the formula proposed by Williams does not fulfill its own criterion because it lacks this critical element.
(d). Partitive word advocates should note that I am not referring to the micro pragmatic and extra linguistic context for evidence for their claim that prototokos conveys an intrinsic partitve property. Rather I am centering precisely on the isolated Greek linguistic component prototokos. In order to answer these questions it is critical to note that the context and micro pragmatic context cannot be used, including the immediately preceding or following grammatical linguistic markers. The reason is because the partitive force or value may actually be leaking over to the term prototokos as an implicature from outside influence. That is, such a semantic partitive force may actually be conveyed from non-linguistic influence (implicature), and not from a linguistically intrinsic partitive property in prototokos itself. To advocates who argue that prototokos is inherently a “partitive word,” other contexts where prototokos also appears (e.g.’ the instances in the LXX) cannot be used to argue their position (that prototokos is intrinsically a partitive word). Why? The reason is simple: Advocates who argue this position first need to determine that it was not the context in all those other instances (anything going beyond the encoded linguistic item at issue) that conveyed that partitive element (implicature) over to the word each and every time it appeared. If that were the case, “prototokos” would not be an intrinsic “partitive word,” linguistically, since it would be the context in all those instances that would allow it to be; that is, “the partitive use of a word” (pragmatics), which is something quite different than “the use of a partitive word” (lexical semantics). It should be noted that the “partitive word” argument does not center on pragmatics or contextual interpretation, but rather on an appeal to the intrinsic properties of the linguistic item itself. The argument is presented as if there were some support from lexical semantics to substantiate this claim. This is a claim that I have requested linguistic evidence for, and a claim for which no linguistic evidence has been produced
(e). Hence, in order for advocates of this view to prove their position, they first need to isolate the linguistic component before analyzing its intrinsic semantic properties. Indeed, this approach is also in accord with Furuli’s own endorsed method of linguistic analysis for translation, where, as in the natural sciences, “a basic principle is to break everything down to the smallest possible units and then study each unit.”[xiii] In fact, for those who argue that prototokos intrinsically conveys an inherent partitive semantic value, this isolating approach is the only procedure which can possibly guarantee adequately isolating outside contamination from pragmatic contextual influences, and this would be theoretically ideal for examination. Of course, the problem is that on the other hand (aside from the scientific problems involving such a feat, see section (c) above), the notion that the word is the minimal unit of communication provides no support for the partitive word view either. Hence, it would seem that the task of attempting to prove that the isolated component prototokos possesses a so-called partitive semantic property intrinsic in its linguistic form is a futile endeavor. From a different angle, if one were to argue that the micro pragmatic context conveys a partitive force (implicature) over to prototokos, then this would simply be begging the question, since that would already assume (without proving) that the entire genitive phrase prototokos pasies kitiseos is a partitive genitive to begin with. Also, as noted already, this would not prove that prototokos conveys a semantically intrinsic partitive force. Now, even granting that a particular word is used in a legitimate partitive genitive construction (even if it occurs over 70 times), I still fail to see how that would demonstrate linguistically that the isolated word itself possessed an intrinsic semantic partitive value. This is clearly confusing lexical semantics with pragmatics. Thus, if there really is pure linguistic evidence that such a partitive semantic property existed intrinsically in prototokos, then it must be demonstrated from the term in isolation. I do not believe that it can be done, and this is the challenge that I pose to advocates of this view.
[i] However, Furuli (1999: 252) notes that Nigel Turner’s comments may understand and describe a partitve sense of prototokos, although this does not lead Turner to conclude that the prototokos was included in a class of created beings. Turner writes, “Might it not be a partitive genitive? ‘Among all created things.’ I would retain the manifest meaning of prototokos ‘firstborn’—but in the sense that the Messiah was said to be firstborn—and interpret the word closely identifying Christ with the family of which he is head, i.e., the whole of creation which looks eagerly for redemption. It has a parallel in the epistle to the Romans where St. Paul again described him as a new Adam, closely identified with believers as an Archetype of a fresh stage or leap forward in the collective evolution of all the creatures of God, in the onward march towards the goal of achieving what Christ is himself—the ‘icon of the invisible God.” (Nigel Turner, Grammatical Insights into the New Testament (Edinburgh, T. & T. Clark, 1965, p. 124). However, from a purely linguistic perspective, no linguistic evidence is provided from the lexis itself to substantiate the notion that prototokos intrinsically conveys a partitive semantic value (see section 3 c).
[ii] Retrieved June 14, 2003 from the World Wide Web:
[iii] For information on the interaction between Dr. Robert Keay and Wes Williams see the below listed web page: Retrieved June 14, 2003 from the World Wide Web:
[iv] Retrieved June 14, 2003 from the World Wide Web:
[v] Retrieved June 14, 2003 from the World Wide Web:
[vi] Although some, for example Peter Kirk, have noticed that there seems to have been entire chunks of information actually lifted from Furuli’s original post and then passed off as William’s own original material and research. This would naturally lead one to question whether or not Williams himself conducted this research which he says he conducted since the words in his post are apparently not his own. Retrieved June 14, 2003 from the World Wide Web:
[vii] In fact, in Nov of 1998 Furuli also stated that the English “firstborn” was “a partitive word”: “There can be little doubt that, if not an important theological question were involved, the genitive would be taken as partitive, because ‘firstborn’ is a ‘partitive’ word. . .” Retrieved June 14, 2003 from the World Wide Web:
[viii] Retrieved June 14, 2003 from the World Wide Web:
[ix] Retrieved June 14, 2003 from the World Wide Web:
[x] Retrieved June 14, 2003 from the World Wide Web:
[xi] Retrieved June 14, 2003 from the World Wide Web:
[xii] Rolf Furuli, The Role of Theology and Bias in Bible Translation with a special look as the New World Translation of Jehovah’s Witnesses (Huntington Beach, Ca: Elihu Books, 1999)