επι - Matt 14:14, 15:32; 4:6, 14:11

[David Lim]

I had previously thought it was quite arbitrary.

Difference in form. Difference in meaning.

Hello Mike! I would indeed like to distinguish different shades of meaning wherever possible, but could not in the following:

[Matt 14:14] και εξελθων ο ιησους ειδεν πολυν οχλον και εσπλαγχνισθη επ αυτοις και εθεραπευσεν τους αρρωστους αυτων
[Matt 15:32] ο δε ιησους προσκαλεσαμενος τους μαθητας αυτου ειπεν σπλαγχνιζομαι επι τον οχλον οτι ηδη ημεραι τρεις προσμενουσιν μοι και ουκ εχουσιν τι φαγωσιν και απολυσαι αυτους νηστεις ου θελω μηποτε εκλυθωσιν εν τη οδω
Do you mind explaining the difference in meaning between the use of "επι" in Matt 14:14 and Matt 15:32? One is with the accusative but the other is with the dative.

[Matt 4:6] και λεγει αυτω ει υιος ει του θεου βαλε σεαυτον κατω γεγραπται γαρ οτι τοις αγγελοις αυτου εντελειται περι σου και επι χειρων αρουσιν σε μηποτε προσκοψης προς λιθον τον ποδα σου
[Matt 14:11] και ηνεχθη η κεφαλη αυτου επι πινακι και εδοθη τω κορασιω και ηνεγκεν τη μητρι αυτης
And what about these two, where one is with the dative but the other is with the genitive? For this in particular LSJ says the following about "επι" used to denote position:

B. WITH DAT.:
I. of Place, upon, just like the gen. (hence Poets use whichever case suits the metre, whereas in Prose the dat. is more freq.)

And I remember that a long time ago I also mentioned the apparent equivalence between "ειπεν προς αυτον" and "ειπεν αυτω" (Luke 7:40,43; 9:50,58 ...). Mark and John hardly use the former. Do you have an explanation for this too?

Thanks!

[David Lim]

I just saw another example:
[Matt 14:25] τεταρτη δε φυλακη της νυκτος ηλθεν προς αυτους περιπατων επι την θαλασσαν
[Matt 14:26] οι δε μαθηται ιδοντες αυτον επι της θαλασσης περιπατουντα εταραχθησαν λεγοντες οτι φαντασμα εστιν και απο του φοβου εκραξαν
This pair is attested by Codex Sinaiticus and Codex Vaticanus. Surprisingly the Byzantine text (Codex Basilensis, Codex Boreelianus) has the order reversed. Either way, the alternation in each of these manuscripts shows that there is no actual difference in meaning. Instead it could be just stylistic variation with no intended distinction.

[Stephen Carlson]

Hi David. I think Mike has been pretty busy for a while and your request about the finer nuances of the preposition ἐπί is something that would take a lot of effort to adequately respond to. Although there are some good, recent studies on the meaning of prepositions in Classical Greek (e.g., Silvia Luraghi, Pietro Bortone), scholarship on their Koine meanings has unfortunately lagged behind.

So your specific question about ἐπί is a pretty big ask. If people want to participate, that would be great, but I think on this question we're going to be limited by the same common set of reference materials--and equally frustrated by them. (LSJ is over a half-century old.)

As to the remark quoted in the original post of this topic, I don't recognize "arbitrary" as a valid linguistic or exegetical explanation. In fact, it is not an explanation at all but a refusal to provide an explanation or worse, a claim that no explanation is possible. For Koine Greek, we have no native speakers to consult and our understanding is naturally limited based on the nature of the surviving evidence. Sometimes, agnosticism is appropriate; sometimes we have to concede that, based on our current understanding of the evidence, we don't know what difference in function (semantic, pragmatic, sociolinguistic, etc.), if any, is meant by a particular difference in form. But admitting our ignorance on a particular point is not the same thing as claiming that it's arbitrary. The latter shuts down the investigation, while the former permits it to go on.

As far as I am aware, "Difference in form. Difference in meaning." is not an iron law of linguistics, but it is a helpful reminder not to give up too soon in one's investigation.

[cwconrad]

Hi David. I think Mike has been pretty busy for a while and your request about the finer nuances of the preposition ἐπί is something that would take a lot of effort to adequately respond to. Although there are some good, recent studies on the meaning of prepositions in Classical Greek (e.g., Silvia Luraghi, Pietro Bortone), scholarship on their Koine meanings has unfortunately lagged behind.

So your specific question about ἐπί is a pretty big ask. If people want to participate, that would be great, but I think on this question we're going to be limited by the same common set of reference materials--and equally frustrated by them. (LSJ is over a half-century old.)

As to the remark quoted in the original post of this topic, I don't recognize "arbitrary" as a valid linguistic or exegetical explanation. In fact, it is not an explanation at all but a refusal to provide an explanation or worse, a claim that no explanation is possible. For Koine Greek, we have no native speakers to consult and our understanding is naturally limited based on the nature of the surviving evidence. Sometimes, agnosticism is appropriate; sometimes we have to concede that, based on our current understanding of the evidence, we don't know what difference in function (semantic, pragmatic, sociolinguistic, etc.), if any, is meant by a particular difference in form. But admitting our ignorance on a particular point is not the same thing as claiming that it's arbitrary. The latter shuts down the investigation, while the former permits it to go on.

As far as I am aware, "Difference in form. Difference in meaning." is not an iron law of linguistics, but it is a helpful reminder not to give up too soon in one's investigation.

It appears to me that what's being touched on here is a metalinguistic question and one which, if it is actually to be discussed, probably belongs in a different sub-forum. I've been troubled by the notion that "choice implies meaning" in such matters as these collocations of prepositions and case-forms of substantives. I'm not sure that a conscious choice of more-or-less equivocal locutions is regularly, consistently being made by all speakers.I certainly am not convinced that there's a fundamental or significant difference in meaning between λέγει αὐτῷ and λέγει πρὸς αὐτόν. With regard to the prepositions and case-forms associated with them, I think there is a weakening or softening of older clear distinctions in Hellenistic Greek and that this weakening or softening is related to the slow and gradual diminution of the dative case in spoken Greek, a process that will take centuries to reach the point where the dative case ceases to be functional in spoken Greek. I'm quite curious about how these processes of linguistic change take place. One factor, I would think, might be the ease of using a preposition that has a more or less fixed sense with an accusative case, whereas originally the prepositions seem to have been adverbs clarifying the usage of the case-ending. Is it easier to say λέγει πρὸς αὐτόν than to say λέγει αὐτῷ? If MG uses ἐπί with an accusative for all senses of ἐπί, when and how did that come about? My guess is that different speakers and writers at the time of the composition of the GNT employed different degrees of precision in their choice of locutions, such that some made their intended meaning quite clear while others wrote relatively vaguely (e.g. the author of Ephesians).

[David Lim]

Hi David. I think Mike has been pretty busy for a while and your request about the finer nuances of the preposition ἐπί is something that would take a lot of effort to adequately respond to. Although there are some good, recent studies on the meaning of prepositions in Classical Greek (e.g., Silvia Luraghi, Pietro Bortone), scholarship on their Koine meanings has unfortunately lagged behind.

So your specific question about ἐπί is a pretty big ask. If people want to participate, that would be great, but I think on this question we're going to be limited by the same common set of reference materials--and equally frustrated by them. (LSJ is over a half-century old.)

Sure, I am in no hurry to get an answer. I just put this out for everyone in case anyone has some insight into any possible difference between the instances of each pair. Until then, I will stick to my working hypothesis that there is none.

As to the remark quoted in the original post of this topic, I don't recognize "arbitrary" as a valid linguistic or exegetical explanation. In fact, it is not an explanation at all but a refusal to provide an explanation or worse, a claim that no explanation is possible. For Koine Greek, we have no native speakers to consult and our understanding is naturally limited based on the nature of the surviving evidence. Sometimes, agnosticism is appropriate; sometimes we have to concede that, based on our current understanding of the evidence, we don't know what difference in function (semantic, pragmatic, sociolinguistic, etc.), if any, is meant by a particular difference in form. But admitting our ignorance on a particular point is not the same thing as claiming that it's arbitrary. The latter shuts down the investigation, while the former permits it to go on.

As far as I am aware, "Difference in form. Difference in meaning." is not an iron law of linguistics, but it is a helpful reminder not to give up too soon in one's investigation.

Actually I think that I am more particular about preserving differences in meaning than most translations, but more often than not, I finally have to give in to the evidence that two different forms are actually identical in function and emphasis, and differ only in length and how they sound. That happened too many times already, especially with prepositions, for me to dismiss them all as my ignorance. For example, I thought that perhaps "λεγει προς αυτον" emphasized a more directed speech but "λεγει αυτω" was the more generic one, however after looking at all the occurrences, I concluded that it was a non-existent distinction. I would be glad to hear if anyone has a plausible suggestion of a distinction, as I can find none, which probably suggests that the large majority of ancient Greek speakers also could not and thus did not distinguish them.

As an example of equivalent phrases in English, not a dead language, "it's" as used in your post is equivalent to "it is". You used "it's" once and "it is" twice. Was there a reason for the difference? I must say that if there was, I was oblivious to it; it escaped my notice even though I am a careful reader, which means that probably no one would have picked up a distinction either, before reading this paragraph. I only found it on a meticulous examination of your post to find an example. "it's" can mean "it has" also, but I don't consider that because no one would have mentally considered that option when reading your post, because the context agrees with the default "it is". Another example is "in order that" which is equivalent to "so that", where "so that" can also mean "with the result that" but it is so rare that it never occurs to the audience unless the default meaning fails to fit the context. All these imply that ultimately there is no distinction between such pairs in the mind of the audience except for the length of the phrase, the time it takes to say it and how the phrase sounds, given the appropriate context (including whether it is formal or informal). Phrases that are truly equivalent in all contexts are harder to find, but I remember seeing some: "with respect to" = "with regard to" = "in regard to". The first two are the most common and differ extremely slightly in Mathematics. The last two are really equivalent though "in regard to" is less frequently used. Another example I just found is "aforesaid" = "aforementioned", with the former less frequent but having exactly the same meaning.

Anyway here is a list of all such pairs involving "επι" I can find in Matthew alone (with sporadic examples from elsewhere):
(I have categorized them according to the semantic domains because other semantic domains seem to be restricted to a specific cases. I have also quoted only the relevant part of each verse.)

on (position)

[Matt 19:28] "εν τη παλιγγενεσια οταν καθιση ο υιος του ανθρωπου επι θρονου δοξης αυτου καθισεσθε και υμεις επι δωδεκα θρονους κρινοντες τας δωδεκα φυλας του ισραηλ"
[Luke 20:30] "και καθισεσθε επι θρονων κρινοντες τας δωδεκα φυλας του ισραηλ"

[Matt 9:9] "ειδεν ανθρωπον καθημενον επι το τελωνιον ματθαιον λεγομενον"
[Matt 23:2] "επι της μωσεως καθεδρας εκαθισαν οι γραμματεις και οι φαρισαιοι"

[Mark 11:2] "ευρησετε πωλον δεδεμενον εφ ον ουδεις ανθρωπων κεκαθικεν"
[Mark 11:7] "και επεβαλον αυτω τα ιματια αυτων και εκαθισεν επ αυτω"

[Rev 14:14] "και ειδον και ιδου νεφελη λευκη και επι την νεφελην καθημενον ομοιον υιω ανθρωπου"
[Rev 14:15] "και αλλος αγγελος εξηλθεν εκ του ναου κραζων εν φωνη μεγαλη τω καθημενω επι της νεφελης"

[Matt 14:25] "απηλθεν προς αυτους ο ιησους περιπατων επι της θαλασσης"
[Matt 14:26] "ιδοντες αυτον οι μαθηται επι την θαλασσαν περιπατουντα"

[Matt 4:6] "επι χειρων αρουσιν σε μηποτε προσκοψης προς λιθον τον ποδα σου"
[Matt 14:11] "ηνεχθη η κεφαλη αυτου επι πινακι και εδοθη τω κορασιω"

on (object of thought)

[Matt 14:14] "ο ιησους ειδεν πολυν οχλον και εσπλαγχνισθη επ αυτοις"
[Matt 15:32] "σπλαγχνιζομαι επι τον οχλον"

over (in authority)

[Matt 24:45] "ο πιστος δουλος και φρονιμος ον κατεστησεν ο κυριος αυτου επι της θεραπειας αυτου"
[Matt 24:47] "επι πασιν τοις υπαρχουσιν αυτου καταστησει αυτον"

[Matt 25:21] "ευ δουλε αγαθε και πιστε επι ολιγα ης πιστος επι πολλων σε καταστησω"

Incidentally, I noticed that "επι την γην" is only used together when there is motion, but this does not hold for other objects, thus I hypothesize that it is necessary to distinguish between "on the earth" (position) and "to the land" (destination) / "on the ground" (direction). For other objects that do not have such compatible multiple meanings, there is no necessity for any distinction and hence both the accusative and genitive are interchangeable when denoting position, as the above examples demonstrate.

And, Carl, is there any reference on the development of the syntax of the preposition that is available online? Or at least with respect to "επι" in the above-mentioned semantic domains? I am curious to know why certain domains are affected but not others, apparently.

[Stephen Carlson]

Sure, I am in no hurry to get an answer. I just put this out for everyone in case anyone has some insight into any possible difference between the instances of each pair. Until then, I will stick to my working hypothesis that there is none.

So your null hypothesis is that there is no difference (of what kind?) for these two forms? OK. But a null hypothesis can only be rejected or failed to be rejected. It cannot be confirmed. That's why I don't understand the objection to admitting one's ignorance in such cases.

[David Lim]

Sure, I am in no hurry to get an answer. I just put this out for everyone in case anyone has some insight into any possible difference between the instances of each pair. Until then, I will stick to my working hypothesis that there is none.

So your null hypothesis is that there is no difference (of what kind?) for these two forms? OK. But a null hypothesis can only be rejected or failed to be rejected. It cannot be confirmed. That's why I don't understand the objection to admitting one's ignorance in such cases.

My current hypothesis is that there is no difference in meaning or emphasis. There would still be a difference in the length of the phrase, the time it takes to say it and the way it sounds. As for why I don't see that it is ignorance, we discussed this at length before but probably can't quite agree soon. My assumption is that people learn a language by generalizing patterns by default and by memorizing exceptions when the patterns don't produce the correct answer. Thus when there is no discernible difference in meaning, no one would learn that two phrases mean different things, but would instead consider them equivalent by default, until coming across situations in which the distinction is made, either explicitly in contrasting the two or implicitly by the conspicuous presence of one and absence of the other.

Back to the preposition in question, I have to either go with the hypothesis that these variations in syntax do not bear any real distinction in meaning, either because of dialectic differences or simply being equivalent (like the English examples I gave), or go with the hypothesis that there is some subtle nuance between the variations in each pair I mentioned that no one has been able to identify. The latter is actually the hypothesis that cannot be confirmed, because I am quite certain that any purported distinction would be so small as to be doubtful. The former, rather, is actually confirmed by the abundance of dialects in many languages, especially those spoken widely. So I indeed expect the same to be true for ancient Greek.

However, as I said earlier, I do really want to discern any nuance possible, so if you or anyone can demonstrate a difference between the syntactical alternatives within any of the semantic domains I listed, I would be glad to know it!

[Stephen Carlson]

Back to the preposition in question, I have to either go with the hypothesis that these variations in syntax do not bear any real distinction in meaning, either because of dialectic differences or simply being equivalent (like the English examples I gave), or go with the hypothesis that there is some subtle nuance between the variations in each pair I mentioned that no one has been able to identify.

You don't "have to" choose either hypothesis. No one is forcing you to do anything. Another possibility is that you can acknowledge that you simply don't have the information to decide between them or even another possibility that you haven't considered. As a general principle, it is better to keep an open mind than to close it prematurely.

[Tony Pope]

Alford discusses the different cases used with ἐπί. I don't have his commentary digitized so can merely point to some pages at archive.org:

Matt 19.28 (he sees a distinction between the one who rightfully sits on the throne and those who are, so to speak, promoted to that position -- but note there are textual variations in the verb) http://archive.org/stream/greektestamen ... 8/mode/2up

Matt 14.25-26 http://archive.org/stream/greektestamen ... 0/mode/2up

Rev 4.2 (long note in which he admits it is hard to account for the variations but observes that at the first mention of the fact of sitting the accus. is used, i.e. introducing the topic is sufficiently like motion towards a point to elicit the accusative) http://archive.org/stream/greektestamen ... 4/mode/2up

Rev 14.14 (refers to his note on Rev 4.2) http://archive.org/stream/greektestamen ... 0/mode/2up

If Alford is right (and no doubt many will think this is far-fetched), there seems to be a tendency to choose the accusative in certain contexts, particularly where the writer feels the necessity to make a subtle distinction. Otherwise apparently the genitive or dative is the default.

[David Lim]

Hello Tony! Thanks for your suggestions. Here are my thoughts on them.

Matt 19.28 (he sees a distinction between the one who rightfully sits on the throne and those who are, so to speak, promoted to that position -- but note there are textual variations in the verb) http://archive.org/stream/greektestamen ... 8/mode/2up

At first I also thought that this might be true, that the accusative connotes a prior "moving or being moved" to that location whereas the genitive simply denotes location. However, as I pointed out, Luke 20:30 has the genitive. If Luke was using the default genitive, it means he did not have such a distinction in mind, so how do we know that Matthew did? I don't assume beforehand that a difference implies intended distinction, so I cannot conclude anything based on the difference in Matthew alone, since he also uses the genitive in Matt 23:2.

Matt 14.25-26 http://archive.org/stream/greektestamen ... 0/mode/2up

I can't buy this explanation because both instances occur with the same verb, and the one having the genitive is used with "απηλθεν προς αυτους ο ιησους". I am aware of textual variants, but it does not nullify the fact that some manuscripts have such a pair in that order, written by Greeks who would easily have recognized if it was indeed a mistake to use the genitive denoting place with a "verb of motion". If it was not a mistake, then I see no reason to reject the explanation that both of the pair are indeed denoting the same thing. Besides, "περιπατειν" by default means "to walk about", not "to walk on the spot".

Rev 4.2 (long note in which he admits it is hard to account for the variations but observes that at the first mention of the fact of sitting the accus. is used, i.e. introducing the topic is sufficiently like motion towards a point to elicit the accusative) http://archive.org/stream/greektestamen ... 4/mode/2up

Rev 14.14 (refers to his note on Rev 4.2) http://archive.org/stream/greektestamen ... 0/mode/2up

I don't buy this either. Here is what he says:

"εκειτο" is wrongly taken by Bengel as importing breadth; and by Hengstb. as representing the resting on the cherubim. But it is St. John's word for mere local position ... the accus. is perhaps not to be pressed; it may be loosely used as equivalent to the gen. or dat. The variations of the case in this expression throughout the book are remarkable, and hardly to be accounted for. ... The only rule that seems to be at all observed is, that always at the first mention of the fact of sitting, the accus. seems to be used, ..., thus bearing a trace of its proper import, that of motion towards, of which the first mention partakes. But the accus. is not confined to the first mention, witness ch. xi. 16, and no rule at all seems to prevail as regards the gen. and dat.

I would say that the variations of the case is the same throughout the NT, making it unremarkable. How can the case have something do with whether it is first mentioned? And it seems weird to try to identify the "proper import" of a preposition in every possible usage of it, since there is really no such thing in any natural language.

If Alford is right (and no doubt many will think this is far-fetched), there seems to be a tendency to choose the accusative in certain contexts, particularly where the writer feels the necessity to make a subtle distinction. Otherwise apparently the genitive or dative is the default.

Actually, for denoting location, the dative is rare, in Matthew at least, which means that the accusative and genitive are the defaults. So if there is a distinction between cases when used for location, it is probably between the dative (denoting mere position of the object with reference to another) and the other two (denoting the broader location of the action or the object). Also, the accusative must be used when denoting the direction of action. This is the best I can identify so far in the NT. And concerning the other two semantic domains I listed, do you also have any suggestion of an explanation for the variations? I shall continue through the NT and see what more I find.