As many of you know

CAI's Continued Misrepresentations of the Phrase
Heos Hou in Matthew 1:25

Click here for an addendum to this article: a surrejoinder to Robert Sungenis' latest response

As many of you know, Robert Sungenis of CAI has been promoting a book that he and one of his assistants are writing as a response to my book, Who Is My Mother? When he first announced his intentions to write the book, I predicted at that time (through an article that was posted on this web site) that we could expect, at best, an inept attempt at a response due to the utter lack of experience and expertise on the part of its primary author. Sungenis assured the world that he personally would be “helping” its primary author with the Greek, and that we should not fear for the quality of the work.

More recently, Sungenis has provided us with a sampling of both the kind and quality of research that is being conducted for this book. He has written an article in the Q&A section of his web site purporting to answer a question regarding the Greek construction heos hou as it relates to Matt 1:25 and the question of whether or not Mary remained a virgin after the birth of Christ. Sungenis' comments are in blue and my responses are in black:

Question: Why can't we use ‘heos hou’ of Genesis 26:13 (LXX) where he "grew until he became very wealthy" which does not have to mean that he ceased to have his wealth increase after he had become wealthy? Yes it does not demand that he DID continue to grow in wealth, but it seems reasonable to argue that the usage here was definitely NOT intending or enforcing the notion that his wealth ceased to further grow after he became wealthy.

Sungenis: You can use them, and they are good examples — all of them. Notwithstanding, let me add some information to the ongoing ‘heos hou’ debate: ‘Heos’ is a relative adverb. ‘Hou’ is a relative pronoun. When used together, ‘heos’ changes to a preposition which governs the relative pronoun. Essentially, this means that hou really has no effect on the meaning of the couplet, the operative word in the couplet being ‘heos’.

If what Sungenis means by “has no effect on the meaning” of the construction is that both forms still retain conjunctive force, then no one can disagree. If instead he means that both forms are used in identical semantic ranges (as I suspect he means), then he is incorrect. In recent years, particularly with the advent of computer-aided research, New Testament scholarship has discovered with increasing frequency that the oft-assumed paradigms and rules for prepositions, conjunctions, and other various Greek constructions no longer hold up under scrutiny. I will expand on this as we proceed; but we need to bear in mind that computer-aided research in biblical studies has really only been around for a little over a decade. GRAMCord was one of the very first fully functional Greek grammatical search programs. I was one of the beta testers of this program while I was a graduate student at Trinity Evangelical Divinity School, where it was heavily promoted and even required in some classes (such as D. A. Carson’s Advanced Greek Grammar class).

Sungenis: 1) ‘heos hou’ is used 17 times in the New Testament (Mt 1:25; 13:33; 14:22; 17:9; 18:34; 26:36; Lk 13:21; 15:8; 22:18; 24:49; Jn 13:38; Ac 21:26; 23:12; 23:14; 23:21; 25:21; 2Pt 1:19) and ‘heos hou’ is used 81 times in the LXX (Greek translation of the OT Hebrew)

This is inaccurate. This is especially odd since Sungenis’ assistant claims to have read my book. If he had, he surely would have known that there are eighty-five instances of this construction in the LXX, not eighty-one. One hopes this is not an example of the supposed inaccuracies (“whammies,” as CAI calls them) that Sungenis’ research assistant claims to have found in my book, Who Is My Mother, and to which he will be responding in his.

Sungenis: 2) ‘heos an’ is used 19 times in the NT (Mt 2:13; 5:18; 5:26; 10:11, 23; 12:20; 16:28; 22:44; 23:39; 24:34) and 95 times in the LXX

Again, inaccurate. The construction “heos an” occurs twenty times in the NT and 105 times in the LXX.

Sungenis: 3) ‘heos otou’ [sic] is used 4 times in the NT (Mt 5:25; Lk 13:8; 22:16; Jn 9:18) and 14 times in the LXX

Once again, inaccurate. The construction “heos hotou” occurs five times in the NT, not four.

Sungenis: 4) ‘heos’(without a couplet [sic]) is used 106 times in the NT and 1564 times in the LXX

As you might expect by now, this is, once again, inaccurate. The number of instances of heos alone—that is, without the particle (what Sungenis oddly calls a “couplet”)—is 104 times in the NT and 1,454 times in the LXX.

Sungenis: According to Burton's Grammar (a popular Greek Grammar used by Protestants) it states the following regarding ‘heos hou’ [sic]: "In the New Testament ‘heos’ is sometimes followed by ‘hou’ or ‘otou’ [sic]. Heos is then a preposition governing the genitive of the relative pronoun, but the phrase ‘heos hou’ or ‘heos otou’ [sic] is in effect a compound conjunction having the same force as the simple ‘heos’. The construction following it is also the same, except that an never occurs after ‘heos hou’ or ‘heos outo’ [sic]." It is clear from this Protestant Greek grammar, that there is no difference between ‘heos’, ‘heos hou’ or ‘heos otou’ [sic]. They all have the same force and the same meaning.

Sungenis appears to think that the really important point here is that Burton is a “Protestant Greek grammar.” It doesn’t seem to matter to him that he has completely misunderstood Burton’s point. Sungenis thinks the word “force” here is to be equated with “nuance” or “meaning,” as though Burton is saying that heos hou has the same “meaning” as heos alone. Far from it. Rather, Burton (as is clear even in the quotation above—which, by the way, is §330 in Burton; Sungenis doesn’t cite the reference) is referring to the part of speech heos is, with or without the particle. Since the particles hou and hotou are genitives, heos technically acts as a preposition that governs the genitive. However—and this is Burton’s point—the construction heos hou or heos hotou retains the same conjunctive “force” that heos has when it occurs by itself. Burton’s point is not that heos hou has the same nuance as heos alone—only that it acts as the same part of speech; namely, a conjunction. If Sungenis had spent less time gloating over the fact that he found a “Protestant” grammar (whatever that is; Greek is non-partisan), and more time attempting to understand his sources, he might have noticed that I included a detailed analysis of heos hou retaining its conjunctive force in my book on Mary.

Now a brief word on grammars. Even if Sungenis and company could find a grammar that lumps all instances of heos and all heos/particle combinations under one semantic umbrella, it means very little. All grammars are general treatments only, and no grammar purports to be an exhaustive study on any word or phrase. I’ve already mentioned computer-aided research above. Grammars are prime examples of where more recent, computer-aided NT scholarship overturns older works that simply did not have the advantages we have today. Even a hard and fast rule such as Colwell’s rule regarding anarthrous predicate nouns has been extensively modified due to more recent, computer-aided scholarship. It would be only a slight exaggeration to say that older rules are being overturned daily by new studies in this area. Sungenis is relying on works that predate the very ability to look at every instance of these Greek constructions, so numerous are they. What’s worse is that such an appeal to Burton—who wrote his grammar over 100 years ago—betrays that Sungenis has not kept up with current NT research methods in Greek grammatical studies. Does he have anything to contribute from this century? Burton—and any other antiquated grammar for that matter—needs to be supplemented by more recent and more extensive Greek studies.

To illustrate my point further, the very first paragraph in that section of Burton’s grammar states the following:

321. “Heos is properly a relative adverb which marks one action as the temporal limit of another action. It does this in two ways, either (a) so that the beginning or simple occurrence of the action of the verb introduced by heos is the limit of the action denoted by the principal verb [such is the case in Matt 1:25], or (b) so that the continuance of the former is the limit of the latter [e.g., John 9:4]. In the former case heos means until, in the latter, while, as long as.”

One wonders why Sungenis didn’t rather cite or refer to this paragraph, since it is really the only paragraph in that section that deals specifically with the meaning of heos. Here Burton states that there are two basic meanings of heos: (a) “until” (which, according to Burton, limits the action of the main clause to the time denoted by the action of the subordinate clause; and (b) “while” (which, according to Burton, shows contemporaneous action). Matt 1:25 falls under Burton’s category (a) “until,” and certainly not category (b) “while, as long as.” Burton doesn’t cite any instance in which heos means “until” and implies continuation of the main clause.

Sungenis could have made a better case for himself by referring to my book, rather than to Burton, since I allow that heos does sometimes imply continuation. As it is, the use of Burton actually militates against Sungenis, providing what amounts to no support for his view of Matt 1:25, and, indeed, some evidence to the contrary. If I were as zealous as Sungenis to find contrived support for my view, I’m certain I could score many points simply by foregoing any mention of the fact the grammars are general treatments only, and are not to be used in the unqualified way Sungenis uses them—namely, without being supplemented by extended studies on specific grammatical constructions. I could have done this as well; but honesty in scholarship compels me to concede a point that might actually help Sungenis’ case, but that Burton does not mention; namely, that heos, when it occurs alone or with the particle an, can indeed be used to show no reference to continuation or termination of the main clause by the action of the subordinate clause. Again, if Sungenis had just read my book, he would have found evidence more favorable to his own position than Burton offers. The fact that he prefers to misread Burton indicates in my mind that he is interested only in an “appearance” of evidence for his view, and not in dealing with his sources in a fair and evenhanded way.

Sungenis: The only other contingency here is the use of ‘heos an’. This is a special case in Greek. When the clause introduced by heos depends on a verb of future time, and refers to a future contingency, it takes the Subjunctive mood with the use of ‘an’, both in classical and New Testament Greek (Mt 5:18).

Sungenis is no doubt relying on Burton’s § 322 for this, but again fails to cite Burton, writing instead as though this were his own thought. However, in the process Sungenis fails to recognize that Burton’s intended observation is one of mere grammatical form rather than meaning. Burton’s intent is simply not what Sungenis hopes to get out of it. That task is left to others who, having a much greater advantage over Burton (who wrote over a century ago), and using more recent, computer-aided research, have spotted grammatical idiosyncrasies that Burton has missed (partly due to the general nature of grammars, and partly due to the inability to find all instances of a construction in days of old). In-depth research on even one grammatical construction can take years—and there are virtually endless possibilities for grammatical constructions. The point is, no one grammarian is going to catch them all—such would be an impossibly daunting task.

As an example of this, when I attended D. A. Carson’s Advanced Greek Grammar class many years ago, I was assigned the grammatical construction “the infinitive + eis + the anarthrous accusative noun.” Carson gave me an A- on the paper and made the paper required reading for the rest of the class, largely because I had uncovered a nuance for that construction that bears on the interpretation of Heb 6:6 (“to renew them to repentance”). But what grammarian of old would have thought—or even have the ability—to compare every instance of the infinitive + eis + the anarthrous accusative noun? The point, again, is that there are endless possibilities for researching grammatical constructions—and most of these are spurred on by a personal exegetical study of a particular biblical text—so it should come as no surprise to anyone that older scholarship, who did not have computer-aided research, would simply not be able to uncover all possible nuances of all possible constructions. Indeed, in many cases, inquiries were (in past times) limited to biblical literature, with little thought that further research on the same construction in all available Hellenistic literature of that era (such a task must have seemed mind-boggling to pre-GramCord grammarians!) might yield even greater nuances of the construction in question. And indeed, this is just what has happened in countless instances in which long-established grammatical “rules” have, in more recent times, easily been overturned. Sungenis writes his article(s) as though he is completely unaware of this fact.

Nearly fifteen years ago, when Carson wrote Exegetical Fallacies, even then he was able to point out many examples where computer-aided research had already begun to overturn earlier established works. One of the fallacies he addresses in his book is the fallacy of semantic obsolescence. This occurs when one finds a word or phrase in an older writing (such as the LXX) and attempts to superimpose the meaning of that word or phrase onto a later writing (such as the NT). It was a common practice in older grammars to find the various nuances of a word in both Testaments and combine them into an acceptable semantic range for that word. In more recent times, scholars have rejected that approach because it commits the fallacy of semantic obsolescence. It is a much more sound practice to determine what a word or phrase means in the era in which it is written. When I attended Doug Moo's Septuagint class during my time at Trinity, I was required to procure a copy of Liddell and Scott's Greek-English Lexicon—even though I already owned the lexicon of BAGD—precisely because the Greek of the LXX is different than the Greek of the NT (something Sungenis does not seem to know). If the word or phrase one is examining is a NT word or phrase, then one begins by looking at its nuances in the NT and the surrounding Hellenistic literature to see how it was used by the normal Greek speaker of the writer’s day. That becomes the semantic range for that word or phrase for that era. If one wants to discover the etymology (evolution) of the word or phrase, then it is valuable to look at how it was used in the LXX and the other literature of that era. Note well, however, that finding instances of the word or phrase in that literature does not constitute a semantic range for that word or phrase in all eras. All it does is show how the etymology of the word or phrase took place.

Again, Sungenis appears to be blissfully unaware of all this. He informs the questioner above that Genesis 26:13 represents a fair example of heos hou to establish the semantic range of heos hou in the NT, and consequently the meaning of Matthew 1:25. Such a statement is grossly irresponsible, and it betrays an inexcusable ignorance of how grammatical studies and exegesis are done. There are indeed some words that are Hebraisms—that is, the consistent meaning of a word or phrase in the LXX may be adopted by the NT writers due to influence by that literature. However, heos hou is clearly not one of them. A true Hebraism is a word with a specialized meaning that permeates the LXX, and is also clearly found in the NT. Not only are there precious few instances of heos hou that bear the meaning Sungenis proposes for Matt 1:25, even in the LXX, but there is not even one clear instance of that meaning for this construction in the NT itself. That automatically disqualifies it as a Hebraism.

The bottom line in determining what heos hou really means—and just as significantly, when it means that—is to examine every instance of the construction in the era in which you want to establish a semantic range. When we do this for heos hou in Matt 1:25, the overwhelming evidence is that the normal Greek speaker of Matthew’s day would have understood Matthew to imply that Mary and Joseph engaged in normal marital relations after the birth of Jesus.

Sungenis: Or, when the clause introduce [sic] by heos depends on a verb of past time and refers to what was at the time of the principal verb conceived of as a future contingency, it takes the Subjunctive mood without an in the New Testament (eg., Mt 18:30).

Sungenis has lifted this right out of § 324 of Burton, but nowhere gives him credit for it, or even indicates that it is not his. More importantly for our purposes, Sungenis continues to cite Burton even though the points are completely irrelevant to the discussion of the meaning of heos hou in Matt 1:25.

Sungenis: Lastly, no determination of the meaning of ‘heos’, ‘heos hou’, ‘heos otou’ [sic] , or heos an can be made without the context of the passage being involved. As ‘heos’ can be used either to terminate or to continue the action of the main verb (as its linguistic equivalent "until" does in English and many other languages, including Hebrew), so does ‘heos hou’, ‘heos otou’ [sic] or ‘heos an’.

If Sungenis is using the word “continue” as Burton uses it, then he fails to make his point. What Burton means by “continue” is that heos can sometimes denote contemporaneous action of the main clause and the subordinate clause, bearing the meaning “while, so long as” (e.g., Matt 14:22; "Jesus made the disciples get into the boat and go on ahead of him to the other side, while [heos hou] he dismissed the crowd"). In such a case (as Burton indicates), the action of the main clause "continues" while the action of the subordinate clause ensues. Burton doesn’t use the word "continue" in connection with a distinction in the nuances of heos when it means “until”—indeed, Burton doesn’t so much as mention that there even is a difference of nuances borne by the word heos when it means "until" (although there certainly is, as Sungenis would be quick to assert, and I would be just as quick to concede); he simply gives the meaning "until" and "while" and leaves it at that, precisely because it is not Burton’s purpose to delineate the various nuances within each general meaning. Due to semantic obsolescence, it is a grammatical fact that there is not even one instance in which heos hou or heos hotou bears a “continuation” nuance (when it means "until") in the NT or in the contemporaneous Hellenistic literature of that era. The semantic range for heos hou in that era excludes the “continuance” nuance that the construction bore in earlier times.

Sungenis: Of the above references, ‘heos’, ‘heos hou’, ‘heos otou’ [sic], and heos an are used a total of 1,900 times in LXX and NT Greek.

Again, inaccurate. Heos in all its forms occurs only 1,710 times in the LXX and the NT; 1,564 times in the LXX and 146 times in the NT.

Sungenis: Although the preponderance of these usages are clearly designed to terminate the action of the main verb, in a significant number of cases, heos and its associated conjunctions is clearly designed to continue the action of the main verb.

In the case of heos hou, there are approximately seven or eight instances out of eighty-five (in the LXX), zero instances out of seventeen (in the NT), and zero instances out of approximately fifty (in all non-biblical literature between 100 B.C. and A.D. 100) that conform to this usage. If one can call that “a significant number,” then I think anything could be included in that category.

Sungenis: The decision on whether heos terminates or continues the action of the main verb depends on several factors, e.g., whether one or the other makes logical sense; agrees with the context; agrees with the grammatical construction of the passage; does not contradict other known facts; etc.

I suspect the clause “does not contradict other known facts” has likely been introduced as a blank check to fill in “as needed” things such as the anachronistic testimony of Jerome vs. Helvidius. I somehow doubt that the “other known facts” will include things like the NT writers’ mention of the adelphos (“brothers”) and adelphe (sisters”) of Jesus—a word that is never used in NT times to refer to biological relatives outside of biological siblings; or the “before” clause of Matt 1:18—to show that the birth of Christ took place before normal marital relations ensued; or the use of the word prototokos (“firstborn”) in Luke 2:7 rather than the word monogenes (“only born”; cf. Luke 7:12; 8:41-42; 9:38).

Be that as it may, the decision on the meaning of heos hou in Matt 1:25 is even simpler than Sungenis suggests. Simply look up every instance of the construction in the NT and see if any of the other instances bears the required meaning. None does. If you are more enterprising, look up all the instances of this construction in the Hellenistic literature of the NT era to see whether any of those instances bears the required meaning. Again, you will find none. The reason you will find none is because the “continuance” nuance for heos hou when it means “until” no longer exists in NT times. Indeed, even in the LXX it occurs only a few times out of eighty-five instances. What that tells us is that the semantic range was fading even then, and the “continuance” nuance was quickly becoming the victim of etymology and semantic obsolescence. By the time we reach the first century B.C., heos hou no longer bears the connotation required by the Roman Catholic interpretation of Matt 1:25. By the time we reach A.D. 50 (the approx. date Matthew wrote his gospel), anyone speaking or writing heos hou intending the “continuance” nuance would sound just as strange to his contemporaries as someone today speaking and writing in King James English would sound to us. It would be one thing to speak King James English when quoting a Bible passage. It would be quite another to speak it as a normal mode of communication. The latter, in essence, is what Sungenis is asking us to believe about Matthew when he writes Matt 1:25.

Sungenis: In conclusion, two things cannot be asserted regarding ‘heos’ and ‘heos hou’: (1) that ‘heos’, and ‘heos hou’, (as well as ‘heos otou’ [sic], ‘heos an’) always terminate the action of the main verb; and (2) that ‘heos’ and ‘heos hou’ are used differently in Greek grammar.

Unfortunately, Sungenis has demonstrated neither of these points. Sungenis’ point (1) is a straw man. I have never asserted that heos “always terminates the action of the main verb.” Nor have I ever asserted that heos hou always terminates the action of the main verb. There are a few instances in the LXX where it clearly does not. All I have ever asserted—and continue to assert—is that heos hou in all the literature of the two centuries surrounding the birth of Christ, when it means “until,” always terminates the action of the main clause. That is an irrefutable fact. If Sungenis had read my work he would have known this. If Sungenis has an example contrary to my proposed usage for this era, let him produce it—he can’t because it doesn’t exist.

Point (2) is based on Sungenis’ misunderstanding of Burton. Burton simply does not address the various nuances of meaning of heos hou. The most he does is point out that it has the same conjunctive force as heos alone. Indeed, if we were to argue from Burton alone, it would be § 321, not § 322 that applies in this case. As we have already shown, Burton’s § 321 leaves us with the impression that all instances of heos (that mean "until") terminate the action of the main verb. If we were to allow Burton’s general comments to win the day, it would be utterly devastating to Sungenis’ view. Honesty, however, compels us to recognize that Burton’s grammar gives general or primary usage only, and does not address the various nuances within the general usage.

Throughout this article I have called attention to several very basic errors in Sungenis’ presentation that are difficult to explain given that the writer claims to know Greek. The errors include:

Heos hou and heos hotou are variant forms of the same particle and are treated as such by grammarians. Yet Sungenis separates them as though they are of two different classes. It is one thing to separate them for the purpose of enumerating their instances; quite another to leave them separated in your conclusions (e.g., Sungenis never demonstrates that he knows heos hou and heos hotou are in a different class than heos an). That leads me to believe that the writer is unfamiliar with the relationships of Greek particles.
Sungenis consistently—in both his quotations of Burton and his own reflection—transliterates heos hotou as heos otou, ignoring the rough breathing mark on hotou. This is an error so basic that it could be spotted by someone with a week’s worth of Greek training. It would be easy to overlook a mere typo; and if that were all it was there would be little need to point it out. But Sungenis’ transliteration is consistently wrong, and he never once gets it right. That leads me to believe that the writer doesn’t know even basic Greek.
Sungenis seems to be unfamiliar with the language of Greek grammar, referring to the grammatical construction as “heos with the couplet,” rather than heos with the particle. The research betrays the work of a novice rather than someone steeped in current NT and Greek studies.
Sungenis is consistently wrong on the number of instances for each construction cited. I know exactly why he is wrong in most cases. It is because he has counted the wrong thing. He has counted the passages where the construction occurs instead of the instances in which they occur. Some passages contain two instances of the same construction; yet this has passed Sungenis’ notice, and in the process, we are again treated to a sampling of the quality of the research in which CAI engages.
Sungenis’ over-reliance on grammars to establish a rule that can be established only by a careful examination of every instance in its context and in its literary era, shows that he has not received sufficient training in exegesis and Greek grammatical research methods to weigh in on this discussion. Perhaps he received the training during his time at Westminster; but his current practice betrays a woeful misunderstanding of the genre of grammatical studies. Grammatical rules are established by usage. When it can be shown that a particular Greek construction always denotes x in a literary timeframe, taking into account all the literature used in that era and finding no exceptions, that establishes a “rule” of usage. For instance, if I am Granville Sharp, and I want to see if there is any significance to the construction of two nouns governed by the same article and separated by kai (“and”), I would simply do a search on that grammatical construction, examine every instance of it, and observe whether there is some common denominator, or pattern of usage that ties them all together. As it turns out, there is indeed a pattern of usage for Granville Sharp’s rule (though not the popularized version of it). That is just what I have done in my research of heos hou in the NT era. This is how grammatical studies are done. They are not done by running to the oldest grammar one can find and carelessly pulling tendentious excerpts from it that end up not even addressing the issue at hand.

My question is, Who conducted the research that underlies this article? Was it Sungenis’ “helping hand,” or was it the work of his under-qualified assistant whose work he promised to check? If it was Sungenis, then I confess I have been giving him much more credit than he deserves, so basic are some of the errors in his article. If it was his assistant, then my predication about his abilities has been realized, and Sungenis is relying uncritically on the research of someone who is in way over his head. Either way, it is a fair indicator of what we can expect when the final version of the Queenship publication is eventually released.

Eric Svendsen, Ph.D.

Addendum: A Brief Surrejoinder to Robert Sungenis' Latest Response

Robert Sungenis has recently responded to the article above, and I have secured the original on my hard drive just in case it changes. In his latest article, Sungenis digs in his heels more deeply—both exegetically and emotionally—and in doing so commits errors even more egregious than in his prior article. In consultation with some trusted colleagues of mine I have decided to forego any response until the publication of Sungenis' book. It appears all I have done by my first article is (1) helped Sungenis in his more obvious errors (such as the correct spelling of the particle hotou; the correct enumeration of these constructions, many of which he still has wrong—and not for the reasons he appears to think), and (2) caused him to dig in his heels on more serious errors that are absolutely fundamental to his position, and that betray even further that he is in way over his head on this issue. His further clarification of his position—not to mention his critique of mine—has resulted in a plethora of new errors that are just as fundamental to his argument as those of his original article. Rather than point out these errors now—and in the process provide even more help for Sungenis' book-writing efforts (surely, the help I've given him thus far already obligates him to include my name on his Acknowledgments page)—I have, at the prompting of my colleagues, decided to wait until the book is published to post my critique of Sungenis' latest response. In the meantime, the critique as it now stands will be made available to a few trusted colleagues only. I apologize to those of you who may have been awaiting a response. I realize that it may be useful to you now, but it will be even more useful after Sungenis has committed his view to the permanency of the printing press. Thanks for your patience and understanding.

Eric Svendsen. Ph.D.
08/05/02