Saturday, October 23, 2004

JB15823-Jas #17: Retracing the Threads 

(15823) Jason BeDuhn[Sat Oct 23, 2004 12:32 am](John 8:58 -- Jason #17)

Dear Rob,

This post retraces certain threads of argument from earlier in our debate, prior to your posts 15-18. It reminds readers of the key issues, reviews our discussion of English word order, goes over the issue surrounding your post 6, and responds once again to your posts 11 and 12. My next messages will systematically respond to your posts 15-18. First, I must correct two mistakes of mine:

(1) When I said that our debate had gone on for three months, I was mistaken. It only felt like three months. It was actually, as you stated, just a week past two months.

(2) I was mistaken in saying that you had accused me of "ad hominem." That phrase was used by someone else on the site. You never said it, and I was mistaken in thinking that you had. I apologize for that mistake. You said I had used "insulting, maligning, offensive, and personal" remarks in the debate. The record will have to stand for itself on whether or not that was true. Let me say again that it was never my intention, and I apologize for any inadvertent offense I may have given you personally.


In my book, Truth in Translation, I identified three defects in popular modern translations of John 8:58: (1) inversion of normal English order of main and dependent clauses of the sentence (all of my sample except the Living Bible), (2) improper rendering of verbal tense within the syntax of the sentence (all of my sample except Living Bible and New World translation), abnormal capitalization of main clause (TEV, NAB, AB). These are "defects" because they do not render the Greek accurately into English that clearly conveys how the syntax of the sentence works together to produce a coherent meaning. Since in other occurrences of grammar and syntax comparable to that found in John 8:58, these translations do render the meaning accurately, there is some unusual pressure upon them in this verse to depart from their normal practice. I identified this pressure as theological bias rooted in the mistaken idea that Jesus is quoting Exodus 3:14 here. I pointed out that this interpretation does violence to the sentence, by disconnecting the main clause from its dependent clause, and that the Greek of the two verses does not correspond in the crucial way necessary to identify the one as a quote of the other. By offering parallel examples, I demonstrated that the grammar and syntax of the verse is properly translated as "I have been since before Abraham was born." Although none of the translations I compared gets this quite right, the Living Bible comes closest, the New World translation next closest, with the others too far from the meaning of the original Greek to be acceptable.

You and I agreed that the odd capitalization found in the TEV, NAB, and AB "tips the hand" of the translators that they are being guided by the assumption of a connection to Exodus 3:14 that is not present in the Greek and was unknown to any early Christian commentator who read both passages in the Greek. (BeDuhn 107-110, Jason post 1; Bowman 121-122, Rob post 5). But we disagree on the issues of word order and verbal tense, and these two subjects have been the dominant themes of our exchange.

The NW translation differs from those preferred by you only in its rendering of the verbal tense. It agrees with them in the inversion of normal English word order, and I have criticized it alongside of them for that. In your post 3, you denied that the word order found in most English translations of this verse was, in my words, "fractured or broken syntax." You argued that English has flexibility to put dependent clauses before main clauses, and this does not constitute "fractured" syntax. I clarified that my characterization referred to both the word order and the disharmony of verbal tenses between the main and dependent clauses, and that these two aberrations together merited that characterization. You subsequently (your post 5) acknowledged that that was my meaning. I went on in my post 4, making use of one of the most respected modern grammars of English, to demonstrate that the noted flexibility in ordering clauses in English is not found in connection with the be-verb. You subsequently (your post 5) acknowledged that it was "unusual" for the be-verb to be employed with the same flexibility of word order found with other verbs. I introduced the difference between adjuncts and complements to the verb, which I had left out of my book as too technical for the broad audience I was addressing. This difference is crucial in determining how flexible one can be with the order of clauses in a sentence. Since the dependent clause in John 8:58 is a verbal complement, not an adjunct, it should not be preposed to the main clause. You subsequently (your post 5) acknowledged my explanation of this rule of English grammar (which you summed up as: "when we use the "be" verb with a predicate complement, that complement follows the "be" verb rather than preceding it. The only exceptions are irrelevant to John 8:58 [e.g., the locative "Here I am" or relative clauses such as "which you are"]") without argument or further comment. You had offered a long list of English passages from the Bible which you felt demonstrated the ability to prepose the dependent clause. I pointed out that most of them did not involve the be-verb, and those that did involved adjuncts, not complements, to the be-verb. These arguments of mine went unanswered in all subsequent posts, and so, unless you have something more to add, can be considered to have prevailed.

I provided an extensive (but far from exhaustive) list of pronoun+be-verb statements in the New Testament to illustrate when and how these were employed. This served to demonstrate that English usually reorders the sequence of words relative to the Greek in acknowledgment of the different demands of syntax between the two languages. It showed how some uses were closer to John 8:58 than others, and how the closer parallels generally avoided the final position of the be-verb in the clause, so that the verbal complement preposed to the main verb in the Greek would follow the main verb in English, since English syntax normally follows this order. This contribution, which amounts to five printed pages, was never acknowledged or responded to by you.

I further argued that English normally employs alternatives to the be-verb when we wish to make absolute existential statements, that the use of the be-verb in this capacity had fallen out of English usage since the days of King James. You initially (your post 5) tried to dispute that English had changed that much in the last three hundred years, but wisely dropped this position as untenable.

In your post 5, you conceded that the order found in most translations of John 8:58 was "unusual or even odd." In my reply I said that this admission was sufficient to place the burden of proof on a defender or "unusual or odd" word order in an English translation. You had suggested that the word order of the original was "also unusual," and this seemed to be the line of argument you intended to follow to defend the unusual or odd English order. But I stated there was nothing at all unusual about the Greek word order, and asked directly:

Please be precise what is it that you consider out of the ordinary for greek grammar here?

You have never replied to this question. I should also repeat here that there is no direct correlation between Greek word order and English word order. In my book, I quoted Orlinsky & Bratcher on this point. They referred to the notion that "faithfulness in translation demands that the word order of the original be reproduced," and commented, " This , of course, is simply wrong" (History of Bible Translation, 1991, page 251), a view that is shared by all the major modern translations. An "unusual" order in one does not dictate "unusual" order in the other. It is only when an "unusual" word order in Greek has some semantic significance that that significance must be rendered in English is some way – by word order if appropriate. But none of this has been demonstrated for John 8:58, by you or anyone else. Greek preposes dependent clauses more freely than English does, and quite regularly has the main verb at the end of the sentence in a way that English does not in sentences like John 8:58 when a verbal complement in present in the sentence, or when the be-verb is involved.

With your post 6 and subsequent posts, you have made no further argument on the issue of word order, leaving my position unrefuted and my refutation of your arguments standing. Thus, unless you wish to mount any new argument, we can consider the issue of word order closed. The main clause should proceed the dependent clause in an accurate translation of John 8:58: "I am/have been before Abraham was born." This correct word order acknowledges the function of the dependent clause "before . . ." as a depictive complement of the main verb, a function that you have accepted without argument. It follows logically and necessarily that the main verb is semantically incomplete in this sentence, and must be read together with its depictive complement to reach the full significance of the verb. Just as "I am" does not fully convey the verbal meaning in the sentence "I am hungry," but needs its complement to communicate meaning, and likewise "I was" does not tell us the complete action intended in the sentence "I was going," so too in John 8:58 the verb only has its full meaning with its complement. Now in Greek the complement has a function of completing the tense property of the main verb. In English, this happens most often with what we call "helping" or "auxiliary verbs." So "I going," while providing full conveyance of the action involved, is not complete as regards the tense of the action: "I am going" is present tense, while "I was going" is a past tense. It is rare in English for the same verbal form to be read one way (say, as a present) and another way (say, as a past) when combined with a modifying word, phrase or clause. The only examples that come to mind are orthographic coincidences, for example:

"I read the Bible." vs. "I read the Bible yesterday."

Someone reading the latter sentence might take "read" as a present tense until observing the adverbial complement "yesterday," which alters the tense of the verbal form. This change in the significance of a verbal form is much more common in Greek, as is recognized in all the principal grammars. But in English a verb usually changes not only significance but form when modified temporally by an adverb, adverbial phrase, or clause. So "I see you," when modified by "yesterday," cannot remain "I see you yesterday," but must become "I saw you yesterday." Or, if "yesterday" is part of a phrase with "since yesterday," it cannot be "I see you since yesterday," but must become "I have been seeing you since yesterday."

All that remains in your avoidance of this conclusion is the claim that "I am" is a "predicate absolute," and hence the "before" clause is an adjunct, not a complement. Now you have stated this claim both in respect to the English and in respect to the Greek, so I will need to address both sides of this claim. I will comment on the Greek in my reply to your post 17. On the English, first of all, it would have to be "I exist," not "I am" to be an absolute and, second, you would have to read the sentence to mean that Jesus is declaring his present existence plain and simple, not his existence in any time reference to Abraham. Since this breaks the sentence up into meaningless and decontextualized fragments, it is unacceptable. In context, Jesus is clearly saying he was already in existence at a particular point of past time, and in English this requires the dependent clause to serve as a verb complement, not an adjunct, the verb to not be read absolutely but to be completed by the sense of the dependent clause, and a resultant shift in the verb from the simple present to the past or past progressive.

WHAT MAKES A PPA? (your post 6 revisited)
Let me show you how ambiguities in the wording of your claims, and false suppositions in your assessment of the grammars, led to false leaps in your conclusions in your post 6 that had the effect of misrepresenting the evidence.

Your third conclusion in your post 6 was that 8 of 15 grammars (so roughly half) "regard the use of such an adverbial as part of the definition of a PPA." "Such an adverbial" in your statement referred to your previous, second conclusion, that "Most of these grammars state that an adverbial expression modifies the present-tense verb" in a PPA construction. Note your wording here, which poses as a limiting, defining description of how PPAs are formed. In your post 8 you objected that I "jumped the gun" in correcting your mistake in counting Brooks & Winbery among these "most," since their comment that " a verb alone is sometimes sufficient" clearly indicates that an adverbial is NOT a defining feature of the PPA for them. If all four of your "summary observations" were stated as defining or descriptive of what a PPA is, you have mistated the Brooks & Winbery position on this second point. If the second conclusion was not meant as defining and descriptive of the PPA, what is its function? I explained my criticism further in my post 9 (Sept. 8), that if you had said "Most of these grammars state than an adverbial expression CAN modify," or "OFTEN modifies," then you would have been above reproach on this point. I made the crucial point that in this second conclusion you were using the broadest possible sense of "adverbial," ONE THAT NECESSARILY INCLUDED ADVERBIAL CLAUSES as well as adverbs and adverbial phrases, since you counted among those supporting this conclusion grammars that explicitly cited adverbial clauses among their examples, as well as those grammars that did not actually use the word "adverbial" in their definition, including Goodwin ("expressions of past time"), Smyth ("expression of past time"), BDF ("temporal expression"), Greenlee ("a specific phrase"), McKay ("expression of past time"), and Wallace ("some sort of temporal indicator"). This broad definition, which includes clauses, is then carried forward to your third conclusion by your expression "such an adverbial." I am with you to this point, and recognize with you thatroughly half of the grammars assume that "such an adverbial" standardly is present in the PPA construct, while the other half indicate or imply that PPAs are sometimes formed without such a grammatically adverbial element in the sentence. SO NOTE THAT HALF THE GRAMMARS THINK THERE CAN BE A PPA WITH NO ADVERBIAL MODIFICATION WHATSOEVER, FOUR OTHERS THINK THE PPA IS FORMED BY MODIFICATION WITH A `EXPRESSION OF PAST TIME' OR `TEMPORAL EXPRESSION,' AND ONLY FOUR GRAMMARS SPECIFY MODIFICATION BY `ADVERBIAL EXPRESSION' (JANNARIS, BURTON), OR `A SPECIFIC PHRASE' (GREENLEE) OR `AN ADVERBIAL PHRASE OR OTHER TIME-INDICATION' (FANNING) AS THE DEFINING FEATURE. This is how the statistical summation of these grammars should have been presented, which does not at all support your belief that the "burden of proof" is on someone who maintains that PPAs can be formed by more than just adverbs or adverbial phrases.

As you went on to apply these second and third conclusions of your post 6 in your argument, you improperly transformed statements in these grammars referring to statistical occurrence –
"often," "usually," "generally" – into value judgments of how more or less "clear-cut" a particular example of a PPA would be. This leap from quantity to quality of cases is illegitimate.

This defective handling of the sources then compounds itself when we move on to the fourth conclusion you made in your post 6:

"By an `adverbial expression' of past time most of these grammars evidently mean an adverb or adverbial phrase."

I dissected the falsity of this conclusion in my posts 7 and 9. Among the critical points I made were:

(1) The expression "these grammars" in "Most of these grammars" must be taken to refer to the same set of 15 discussed to this point. You failed to inform anyone that you had actually reduced the number of grammars considered for this point.

(2) "Most" is misleading because only 2 of the 15 grammars even use the phrase `adverbial expression' that you take as a standard. 3 of 15 mention an "adverb" (Robertson, Dana & Mantey, Brooks & Winbery), but not as necessarily the modifying element of a PPA. 2 of 15 use "adverbial phrase" (Fanning, Wallace), but one of these (Wallace) only as an example of the more generic category of PPA modifiers, `some sort of temporal indicator,' which is clearly the broader category Fanning has in mind, too, when he adds `or other time-indication.'

Thus the whole question was raised what you possibly could mean by "most," and the whole claim proves to be baseless. When I first pointed out problems with this posting of yours, you said I should "read my four points together and understand their logical sequence and relationship." Therefore, the fourth conclusion is meant, in sequence, to narrow how "adverbial" is understood, after you have, in conclusion 2, swept under that rubric the broadest possible expressions found in the grammars, most of which quite clearly mean something broader than you attempt here to make them mean. But even if we separate your fourth conclusion from any progressive scheme, and judge it on its own merits, "most" is simply inaccurate: "most" of the grammars do not even use the term "adverb" or "adverbial," so your claim is done in on that fact alone. Moreover,

(3) Of the minority seven who do talk in these terms, Fanning cites two clausal PPAs (Acts 27:33 and 2 Peter 3:4), Wallace one (Acts 27:33), and Robertson one (2 Peter 3:4). This leaves only 4 of 15 grammars that you might construe as "evidently meaning an adverb or adverbial phrase."

(4) Relying on what few examples they happen to cite is a dubious basis on which to claim an absolute exclusion, even among these 4, of clausal cases. What they cite establishes inclusion, but cannot establish exclusion.

You protested (in post 8) that your fourth conclusion "has to do with what the grammars mean by `adverbial expression' or whatever term they use." Pardon me, but you did not say "or whatever term they use" in your post 6, nor does this objection deal with my demonstration that "most" do not mean what you say they mean, nor does it explain while after claiming what "most" mean, you cited only five. Those reading your post 6 "in sequence," as you suggested in your post 8 was the correct way to do so, would have understood you to be basing your "summary observations" on the material you had quoted before hand. That is what a "summary" normally is. Each of your four conclusions was based in the language used within the definition of the PPA by the grammars. I critiqued them on this same basis. In your response in your post 8 you justified your conclusions on the basis of the examples they cite, and pointed back to how you had referred in one sentence in the fourth conclusion to "the examples that the grammars cite, as we shall see." You therefore claimed "my argument was that most of the grammars appear to refer to adverbs or adverbial phrases, because most of the examples they gave were in fact adverbs or adverbial phrases." But not only had I already noted this very sentence in which you looked forward to the examples, but I had already pointed out that what the grammars happen to cite in selecting a handful of examples is far from being adequate to limit the meaning of their definition, which in so many cases is quite clearly worded in very broad and cautious terms. In a subsequent post you acknowledged my point, while still maintaining that such sets of examples created the presumption, if not the conclusion, of such a limitation. This is fallacious. What they cite can inform us what they include, but it cannot inform us what they exclude, because they are each quite clearly citing only a few examples to illustrate the construction they are talking about. Since simple adverbs and adverbial phrases easily
outnumber more complex adverbial clause modification in our sources, the few examples the grammarians pick are statistically likely to favor adverb and adverbial phrase examples.

(5) I had already pointed out your error in stating that "The only grammars that evidently include whole clauses are BDF and McKay (and only because they count John 8:58 as a PPA)," since Winer and Turner also cite it. You rejected this correction, not by disputing that they cite John 8:58, which indeed they do, but by suggesting that since "Winer and Turner say nothing at all about expressions of past time accompanying the PPA verb," they do not count in your fourth conclusion. I don't know anyone reading your fourth conclusion who would have assumed Winer and Turner were excluded, and it runs against your stated claim in your fourth conclusion to be making a summary of the evidence of "these grammars" to leave some of them. Moreover, Turner in fact references Burton for further details on defining the PPA, and so must be understood as affirming Burton's use of "adverbial expression," so Turner needs to be counted with the rest even on your revised criteria. In these five critical ways your fourth conclusion in post 6 is not an accurate representation of the grammars, and must be set aside as void.

Turning to your analysis of the examples of PPAs from the grammars, I showed how problematic your distinction of "contested" and "uncontested" was. For one thing, in only a couple cases were they actually contested as not a PPA by one or more of the grammarians (John 8:58 by Robertson and Wallace; 2 Cor. 12:9 by Fanning). But you extended "contested" status to any verse cited by a grammarian under a different construct than the PPA. I pointed out the varying boundaries and subcategories involved from one grammarian to another that made this less than a "contesting" of the verses' PPA status. And in your latest posts you seem to show agreement on the fluidity of some of these boundaries. This issue was involved in my criticism of you accepting as uncontested a verse that appeared to be contested, because it was cited as a "static present" as well as a PPA: John 15:27. The twist here was that it was cited under both by the same grammar (Dana & Mantey). I pointed out that if the same grammarians could cite a verse under two distinct categories, and you could consider that not contesting (in your post 8 you maintained that it was confusing but not contesting for a single grammar to cite the same verse under two headings), then you should not consider it contesting if two different grammarians cites a verse under two distinct categories respectively. As I said, you know seem to be approaching sharing this viewpoint with me, although you have never directly responded to the particular point of John 15:27. I decided to pursue the question further, however, and found that Dana & Mantey were actually referencing different clauses of the same verse in the respective citations (more on this in a future post). This may have implications for some of the other supposedly "contested" examples, but I have not yet had a chance to look into that. Anyway, it turns out that you were right to consider John 15:27 as not really "contested" by your use of the term, although not for a reason you were aware of at the time. At the same time, you did count as contested a verse that had never been contested: Acts 26:31, which is commented on by three grammars, but in the two of them I have been able to check the comments you cite are not meant to contest or question, but rather explain why it is a PPA despite lacking any adverbial modification of the verb). Furthermore, you decided to overlook the "contested" status of 2 Peter 3:4 and 1 John 3:8 in your post 6, without providing any justification, and I rightly questioned this move. I said you should either include all of the
"contested" examples or none of them, but it was inadmissable for you to accept only those that you thought supported your position and only exclude those that did not. These inconsistencies had the effect of skewing the examples in favor of the position for which you were arguing. Then, following my demonstration that Acts 27:33 was a clausally-modified PPA, you rejected it from the PPA category, though you had accepted it as a PPA when you thought it fit your argument. This was a move that really ran the risk of being misconstrued as recutting the body of evidence to suit your conclusions, especially compared to my rejection of 2 Cor. 12:9 as a PPA, despite the fact that it was one of only four examples of a clause construction among your sample. I did so because that was the truth of the matter, that it was not really a PPA, no matter how much its inclusion would have boosted my position. So you need to be more careful about making these unjustified adjustments of the pool of samples when they become inconvenient for what you are sure must be true.

Any future argument you wish to make concerning what makes a PPA must take account of the following two points of information:

(1) 4 OF 15 GRAMMARS CITE EXAMPLES THAT HAVE NO ADVERBIAL MODIFIER AT ALL. (Luke 2:48 Turner, Moule; Acts 26:31 Turner, BDF, Winer)

(2) 7 OF 15 GRAMMARS CITE EXAMPLES EMPLOYING AN ADVERBIAL CLAUSE (John 8:58 Turner, BDF, Winer, McKay; Acts 27:33 Fanning, Wallace; 2 Peter 3:4 Turner, Winer, Fanning, Robertson) – see my post 7.

Given the fact that adverbs and adverbial phrases are used much more commonly than adverbial clauses (a point you accepted in your post 8), and that identifying a PPA with no adverbial expression is likely to occur only in cases where no other reading is possible, these statistics are quite significant.

Since you have gone on to base your subsequent arguments for a narrower definition of the PPA on the conclusions of your post 6, all of this subsequent argument is voided. The grammars are overwhelmingly against you on this line of argument, and your use of them has been without merit as an accurate representation of how they view the PPA construct. Therefore there is no burden of proof against the PPA understanding of John 8:58, but rather a "level playing field" on which my arguments for a PPA reading, and yours for a gnomic reading can be judged for their cogency, coherence, and supporting comparative examples.

WHAT MAKES A PPA, PART 2 (reply to your posts 11-13) You complain in your post 13 that I have not responded to the bulk of your arguments in your posts 11 and 12. You will have to go back and specify which arguments you feel have not been responded to, because I don't see anything substantial that has not been dealt with. I could be wrong, of course. You also continue to claim that I was not fair on your first conclusion in your post 11, where I pointed out errors in how you assigned points to the various grammars for how limiting they were in their characterization of the modifying elements in PPAs, errors that skewed the results in the direction you prefer them to have. You gave Goodwin a 6, I said it should be a 2. In your post 12, you said it should be a 5. This dispute over scoring the grammars on the conclusions you wished to draw comes about from two distinct flaws in your handling of them: (1) ambiguity in the wording of your points, (2) lack of objective assessment of how grammars fit your devised scale. It is true that I was very brief in my critique, and that might have resulted in my point not being made. Allow me to go over it again more carefully.

In your post 11, you said:
"I assign 6 points if the grammar specifies that THE marker [of a PPA] is an adverb or phrase."

Note the limiting nature of this category. Not that an adverb or phrase is A marker, but THE marker of a PPA. Dana & Mantey specify an adverb, but not as THE marker, and by saying that an adverb is only "generally" associated with a PPA, leaves unclear how far down your score they would slip by what other sorts of modification they would recognize. So rather than speculate on that, I allowed them to stand in score 6. Note also that you could not say, and so avoided saying, ADVERBIAL phrase, though that is what you understood and want us to understand, because one of your four grammars assigned to this score (Greenlee) says "a specific phrase expressing the past aspect." Since not all phrases expressing past time are adverbial (e.g., "a fourteenth day"), Greenlee does not really fit this score either. If you stop to think about it, you will realize that even you cannot take his mention of a "phrase" in its narrow meaning here, since because he says such a phrase is "required," such an understanding of "phrase" would have Greenlee exclude simple adverbs. We'll get to the other two to whom you assign a score of 6 below.

"5 points if it so specifies [that THE marker of a PPA is an adverb or phrase] but allows for other types of temporal expressions."

After initially giving Goodwin a score of 6, you retreated to this score after I pointed out that Goodwin says that the PPA is formed when the present is "used with expressions denoting past time, especially PALAI." Thus he clearly "allows for other types of temporal expressions" (compare Goodwin's Grammar, 1892, sect. 1258:
"The present with PALAI or any other expression of past time . ..").
But due to problems with your scale, his fit at this score is not exclusive, as I will show below. I allowed this score to Brooks/Winbery, to whom you had given a 6, assuming that "THE marker" was overstating what you meant to claim, since they say "An adverb of time is ofte used . . . but a verb alone is sometimes sufficient." In your reply, you insisted that their one example of the latter is not in your opinion valid, and therefore, in effect, their description of the past-time indicator was to be edited to remove the reference to "a verb alone is sometimes sufficient." But the heading under which you were assigning scores was "Description of the Past-Time Indicator," not "Description . . . adjusted according to strength of examples." So your defense of your scoring of Brooks/Winbery is invalid. I said they could be given "at most" a score of 5. I will show below why I put it this way.

"4 points if it describes the marker as an adverbial expression."

This description actually steps back above your score of 5, because in the latter you accepted the broader language of "temporal expression," and here you use the more specific language of "adverbial expression." This confusion in your terminology is problematic for your scale. Notice that Brooks/Winbery's acceptance of cases without an adverbial expression puts them lower than this.
"2 points if it simply refers to the marker as an expression."

Notice that you skipped the score of 3. Note also that Goodwin says "expressions denoting past time" (not "adverbial expression," not "adverbs, but also . . .") of which he gives an example, PALAI (compare his Grammar: "PALAI or any other expression of past time").

So he "specifies" PALAI as an example of an "expression," not as an example of an adverb. That is why 2 is the appropriate score. Since Brooks/Winbery say "a verb alone is sometimes sufficient," they could conceivably slip down your scale even farther than this.

You were outraged that after showing how your assignment of a score of 6 to the above four grammars was inaccurate, I had added "And so forth, throughout your quantifying exercise." You said in your post 12:

"You go on to imply that my entire analysis is flawed."

It is flawed, as I more than implied by going on to show how arbitrary your point scales were in each of your categories of analysis, and how contrived was the correlation between unrelated points (breadth of definition vs, happenstance of discussing translation). But you continue:

"My analysis involved assigning 75 distinct scores (to 15 grammars in 5 categories), and as far as I can tell you have exposed a mistake of 1 point out of 6 in one of those 75 scores."

Impressive statistics, and they fit your claim that my criticism was a "tempest in a teapot," and merited no more than an "oops!" Well, as I have just shown, your accounting of how much error there was in your scoring of just the top score of the first point of 5 is wrong, and I could have criticized it even more. But you are right that I was too quick for you to be informed, rather than just chastised, concerning the problematic nature of your scoring. So what did I mean by "and so
forth"? Since I've covered your top four (that you scored 6), let's just continue down the scoring of this first point of analysis to see what I meant:

"5 points if it so specifies [that THE marker of a PPA is an adverb or phrase] but allows for other types of temporal expressions. 4points if it describes the marker as an adverbial expression."

You assign a score of 5 to Wallace, Jannaris, and Fanning. Wallace says "usually has some sort of temporal indicator, such as an adverbial phrase." Here again, the adverbial phrase is not specified as THE marker of the PPA; it is mentioned as an example of "some sort of temporal indicator," which combined with the given example, makes the most generous scoring a 4; a case could be pressed for "some sort of temporal indicator" to be a 2 or a 0 considering its very broad
wording. Jannaris says that the present "often stands with adverbial expressions denoting past time, such as PALAI." This is clearly, word-for-word, a score of 4. PALAI is given as an example of an "adverbial expression," and not in any limiting sense. Fanning seems to be correctly scored as a 5, Burton correctly as a 4.

"2 points if it simply refers to the marker as an expression."

BDF, Young, Smyth, and McKay are assigned this score correctly. Winer and Turner are scored at 0, as they must be.

So we can see that the "and so forth" had some substance behind it, which I refrained from detailing at first. All of your errors take the form of overscoring, rather than underscoring, and they all occur in assigning the highest two scores (those closest to your view). So there is a clear tendency in these errors, which is to skew the testimony of the grammars in such a way that your position appears more broadly supported than it actually is, and mine seems less supported than it actually is. Your original scoring had 7 out of 15 grammars in the two highest scores, compared to 8 of 15 in the lower three scores. A more accurate accounting (even allowing Greenlee's reference to a "phrase expressing the past time aspect" to be scored high) has 3 or 4 out of 15 grammars (depending on where Brooks/Winbery end up) in the top two scores, and 11 or 12 out of 15 in the lower three scores. So now we can see how the grammatical evidence actually breaks on this issue, and can see how the so-called "broad" definition of the PPA is by far the majority view of it among your set of grammars. Any "burden of proof," therefore, is on a
"narrow" definition of the PPA modification, that is, yours.

I think I have already sufficiently commented on the happenstance nature of grammars commenting on translation, and effectively shown that there is no exclusive correlation between those that do and the "narrow" definition of the PPA.

In my next post, I will turn to your new postings.

Best wishes,
Jason B.

Comments: Post a Comment

This page is powered by Blogger. Isn't yours?