[B-Greek] TARASSW/TARASSOMAI subject-agent affected

Carl Conrad cwconrad2 at mac.com
Mon Feb 25 09:44:06 EST 2008


On Feb 23, 2008, at 4:10 AM, Iver Larsen wrote:

> Comments interspersed below:
>
> ----- Original Message ----- From: "Carl Conrad" <cwconrad2 at mac.com>
> To: "Iver Larsen" <iver_larsen at sil.org>
>>
>>> My own impression is that this reflexive usage of TARASSEIN  
>>> hEAUTON  is
>>> exceptional, although perfectly intelligible: the middle is the   
>>> normal
>>> usage. Of 17 instances of the verb in John's gospel, 6 are active ,
>>> the remaining 11 are middle-passive. Of the 6 active forms all but
>>> that in John 11:33 have external objects.
>>
>> I would prefer to look at this from a semantic viewpoint.
>>
>> TARASSW is basically a divalent verb with an agent or cause as   
>> subject and a patient as object.
>> Somebody or something stirs/upsets somebody or something. If the   
>> subject is a person, I would call
>> it agent, but if it is an event, I would call it cause. In most   
>> cases in the NT the patient is a
>> person (or the heart/spirit of a person), but in John 5:4 and 5:7  
>> it  is water.
>
> I'm not surprised to see this view of the matter expressed by Iver; I
> continue to believe that TARASSOMAI is the basic form and that it is
> intransitive, not passive, and furthermore that TARASSW, the active
> form, is a secondary transitive causative form of TARASSOMAI. I don't
> understand why, when Jesus says, (John 12:27) Νῦν ἡ ψυχή
> μου τετάρακται, ... (NUN hH YUCH MOU TETARAKTAI ... ),  
> we
> are supposed to imagine that Jesus in his reaction here is made to
> react by some external agent. I suppose that FOBEOMAI would be treated
> the same way; I certainly consider it a middle intransitive -- or
> transitive when it takes an accusative object of the fear, while FOBEW
> is the far less common causative active form meaning "frighten."
> ---------------------------------------------------
>
> The difference in our approaches is that yours is syntactic only,  
> while mine is both syntactive and semantic.

I'm not sure whether you use "syntactive" to mean something different  
from "syntactic" or not. While I've generally found you to be more  
consistent in your uses of terminology than some linguists, a switch  
like this confuses me; when one trained as a philologist approaches  
the linguistic(s) tower of Babel, the confusing sounds are rather  
threatening. I will assume, for the time being, that "syntactive" does  
mean "syntactic."

I continue to explore how these frameworks of functional grammatical  
analysis (I don't know whether or not that terminology is correct) may  
be used to describe ancient Greek voice usage adequately. I wonder,  
for instance, how one handles clauses with PASCW such as the following:

Matt. 17:12 ... οὕτως καὶ ὁ υἱὸς τοῦ  
ἀνθρώπου μέλλει πάσχειν ὑπ᾿ αὐτῶν.
  ... hOUTWS KAI hO hUIOS TOU ANQRWPOU MELLEI PASCEIN hUP' AUTWN.

1Th. 2:14 ... τὰ αὐτὰ ἐπάθετε καὶ ὑμεῖς  
ὑπὸ τῶν ἰδίων συμφυλετῶν καθὼς καὶ  
αὐτοὶ ὑπὸ τῶν Ἰουδαίων, ...
... TA AUTA EPAQETE KAI hUMEIS hUPO TWN IDIWN SUMFULETWN KAQWS KAI  
AUTOI hUPO TWN IOUDAIWN, ...

Do we call these forms of PASCW used with an agent constructions  
something like "active transformations of a passive verb"? I wouldn't  
think so.

I need to re-read Halliday's _Intro to Functional Grammar_. I wish  
there were some comparable account of Functional Grammar for ancient  
Greek. Is there anything other than Helma Dik's work on word-order?

The problem, as I see it, is that Greek really has only two voice  
morphoparadigms: "active" and "middle"; the so-called "passive" forms  
are not differentiated from the middle forms in usage, while all of  
the "middle" forms, including the QH forms in aorist and future, may  
be interpreted as middle or passive semantically. Moreover, the  
"active" forms are by no means necessarily active -- I'm convinced  
that "actives" are the default "unmarked" form of the verb, while the  
middle forms (including the passive) are all "marked" for subject- 
affectedness. I know that Iver doesn't like the term "subject- 
affectedness" -- perhaps because "subject" is a syntactic term --, but  
I think it works better than anything else.


> FOBEOMAI is a diffrerent cup of tea, semantically speaking. It is  
> again basically a divalent verb but with a rather different semantic  
> frame. The nucleus is a cognitive verb "be afraid of" with the first  
> semantic role an Experiencer and the second semantic role Source.  
> The first role is syntactically expressed as subject and the source  
> as object or occasionally as a prepositional phrase, e.g.
>
> Mat 21:26: FOBOUMEQA TON OCLON (We are afraid of the crowd)
>
> Mat 10:28a MH FOBEISQE APO TWN APOKTENNONTWN TO SOMA
> (Do not be afraid of those who kill the body)
> Mat 10:28b FOBEISQE DE MALLON TON DUNAMENON KAI YUCHN KAI SOMA  
> APOLESAI
> (but rather be afraid of him who can destroy both soul and body)
>
> The way to find the basic semantic frame is to look at all actual  
> occurrences in the corpus (FOBEOMAI occurs 95 times in the NT, never  
> in the active form.) The next step is to determine the semantic  
> roles of the subject, object and indirect object if they exist. When  
> the first role is Experiencer or Patient rather than Agent, the norm  
> in Greek is that you have a middle verb as you do with FOBEOMAI.
> It is possible to derive from this verb a separate semantic verb by  
> valence addition so that you get A caused B to be afraid of C. This  
> occurs for instance in 2 Ch 32:18 and corresponds to the different  
> verb "frighten". It just so happens that Greek can do this valence  
> addition by using an active form of a basic middle verb. FOBEW and  
> FOBEOMAI ought to have two different entries in a dictionary because  
> of their different semantic frames, just like "fear" and "frighten"  
> are separate entries in an English dictionary.
> I don't think it is particularly helpful to describe FOBEOMAI as  
> either transitive or intransitive, but if you insist, then it is  
> transitive as Matt 21:26 indicates: somebody fears somebody. It is  
> much more helpful to know that the subject is the Experiencer and  
> the source - if it is explicit - is either in the accusative or a  
> prepositional phrase. Very often the source is not explicitly  
> mentioned in the syntax, but that does not make the verb  
> intransitive. An intransitive verb cannot take an object, a  
> transitive verb may occur without a syntactically explicit object.
>
> The main reason that I prefer to analyze TARASSW as different from  
> FOBEOMAI is their different semantic case frames, and this is  
> something which is not clear from a purely syntactical point of  
> view. TARASSW like FOBEOMAI is basically a divalent verb  
> (transitive). The first semantic role in TARASSW is Agent (or Cause  
> when it refers to a concept like an event that has happened or may  
> happen). The second role is Patient. The normal passive  
> transformation is when the Patient is expressed in the syntax as  
> subject. Compare for example John 5:3: AGGELOS ... ETARASSEN TO  
> hUDWR (An angel stirred the water)  with John 5:7 hOTAN TARAXQHi TO  
> hUDWR (When the water is stirred (by the angel)). In the first  
> instance AGGELOS is Agent and subject, while hUDWR is Patient and  
> object. In the second instance hUDWR is Patient and subject while  
> the agent is suppressed/implicit as in most passive transformations.
> One can only see this if one looks at many actual occurrences in the  
> data corpus. It is on the basis of such empirical study that I come  
> to the conclusion that there must have been some event or situation  
> that has deeply upset Jesus in John 12:27. I think it is his  
> upcoming suffering and death or rather the question of whether he  
> must go through with it or whether there is a way out. In this case  
> we don't have a person as Agent, but a concept as Cause. And this is  
> where the results of our two approaches are not so different after  
> all. You prefer to take the MP form as basic and the active as a  
> causative derivation. I take the active for this verb as basic and  
> the passive form as a passive derivation.

I'm a little bit surprised to find FOBEOMAI termed a "cognitive" verb.  
Suzanne Kemmer calls it an "emotion" verb and Allan calls it a "mental  
process middle" -- one that "involves an animate subject that  
experiences a mental affectedness." Allan goes on to say (pp. 64-5),  
"This affectedness can be of an emotional kind, as in φοβοῦμαι  
[FOBOUMAI] 'fear', or of a cognitive kind, as in μιμνήσκομαι  
[MIMNHSKOMAI] 'remember', ἐπίσταμαι [EPISTAMAI] 'know." In  
the present stem, these middle verbs designate states, that is to say,  
they do not involve change over time. However, mental states are  
typically temporary As such, mental states are different from states  
that have a more permanent character such as 'be king', 'be small',  
'be red'. The aorist forms of these verbs generally designate that the  
subject _entered_ a mental state (ingressive meaning of the aorist).  
For example, aorist ἐφιβήθην [EFOBHQHN] 'I became afraid' vs.  
φοβέομαι [FOBEOMAI] 'I am afraid. ... The mental affectedness  
may be caused by an external _stimulus_. This stimulus-participant can  
have a genitive, dative and an accusative case. ... "

I really am not quite convinced that the verb  FOBOUMAI is different  
in kind from TARASSOMAI. There is an active form FOBEW; it doesn't  
appear in the GNT but it does appear in the LXX and it is not so  
infrequent in extra-biblical narrative, especially of military  
engagements (this is where my frequent observation about the  
inadequacy of the GNT as a corpus representative of Hellenistic Greek  
usage come into play). It is pretty clear that the middle-passive  
forms of FOBEOMAI are far more frequent than are the active forms  
(FOBEW). But the same is also true of TARASSW (GNT 6 Active, 11 Middle- 
passive; LXX (24 Active, 94 Middle-passive). I remain convinced that  
both verbs are essentially middle verbs that have not very common  
causative active secondary forms -- and that is why I am loath to  
accede to your view of the "semantic frame" of TARASSW (I wonder, in  
fact, whether the GNT is a sufficient textual corpus for discerning  
the framework of Hellenistic verbal usage). Rutger Allan (_The Middle  
Voice in Ancient Greek: A Study in Polysemy_) lists both FOBOUMAI and  
TARATTOMAI in the category of "mental process middles" which are  
closely akin to "spontaneous process middles. At any rate, let me cite  
Allan on the ambivalence of TARATTW/TARATTOMAI (Ionic-Hellenistic  
TARASSW/TARASSOMAI):

from chap. 2, "The Middle Voice as a Complex Category," pp. 67-68.

	"The mental process middle is related to the passive midde. In both  
types, the subject passively undergoes the event. The difference  
between the two middle types relates to whether or not the event is  
initiated by an external agent. In many contexts, however, the  
initiating entity is irrelevant and remains unexpressed. In those  
cases one cannot decide whether the middle verb is to be assessed as a  
true passive dle. This can be exemplified by:

	(9) Ταῦτα εἰποῦσα, (...), ἐκείνη μὲν  
ἀπλλάγη,
	ἐγὼ δ’ εὐθέως ἐταραττόμην, καὶ  
πάντα μου
	εἰς τὴν γνώμην εἰσῄει
	[TAUTA EIPOUSA, (...), EKEINH MEN APHLLAGH,
	EGW D' EUQEWS ETARATTOMHN, KAI PANTA
	MOU EIS THN GNWMHN EISHiEI] (Lys. 1.17)
	With these words, (...), she  took herself off; I was
	at once perturbed; all that had happened came into
	my mind, (...).

"The middle ἐταραττόμην [ETARATTOMHN] could both be  
interpreted as a true passive (the woman being the external agent), or  
as a spontaneous mental process. The only relevant thing is that the  
subject experiences a mental affectedness, and the initiating entity  
(be it an agaent or a cause) is backgrounded. In such cases it can,  
however, be assumed that the interpretation that is activated first  
and foremost is the mental process interpretation. This assumption is  
based on the fact that the mental process use is far more frequent --  
and is therefore more entrenched -- than the passive use. The  
intransitive meaning thus unctions as a default-interpretation, only  
overruled when explicit indications of a true passive meaning are  
present in the context. This is a form of processing economy: 'if the  
most typical and frequent interpretation makes sense, do not look any  
further.'"
=====

I might add that in footnote 100, p. 68, Allan notes, "Middle verbs  
that designate spontaneous mental and physical changes are subsumed  
under the -_pseudo-passive_ use by Rijksbaron (2002-3: 151-2)."  
Interesting term, is it not, "pseudo-passive"?

I am still somewhat puzzled about a passive SEMANTIC interpretation of  
supposed passive IMPERATIVES, such as:

Acts 2:38 ... καὶ βαπτισθήτω ἕκαστος ὑμῶν  
BAPTISQHTW hEKAASTOS hUMWN

or the strange instance of the middle imperative in Acts 22:16 which  
Iver wants to interpret as a semantic passive:

... ἀναστὰς βάπτισαι καὶ ἀπόλουσαι  
τὰς ἁμαρτίας σου ...
  ... ANASTAS BAPTISAI KAI APOLOUSAI TAS hAMARTIAS SOU ...

In John 14:27 we have the "passive" imperative suggesting that the  
troubling TARACH) of their hearts might be something the disciples  
could check voluntarily.

John 14:27 .... μὴ ταρασσέσθω ὑμῶν ἡ  
καρδία μηδὲ δειλιάτω.
... MH TARASSESQW hUMWN hH KARDIA MHDE DEILIATW.

Admittedly it's rather difficult to instill courage into those who are  
anxious, but two dominant philosophic schools of the Hellenistic era  
with curiously similar conceptions of a "fortress" mentality taught  
lifestyle orientations built a curiously similar emphasis upon  
emotional self-control. The Stoic ideal state is APAQEIA  
(ἀπάθεια) -- not quite what English "apathy" means (Seneca  
suggests it is something like emotional tranquillity). Epictetus  
nicely indicates what's involved in the opening section of the  
Enchiridion: there are things that are outside of our control (τὰ  
μὴ ἐφ’ ἡμῖν TA MH EF' hHMIN) AND things that are under  
our control (τὰ ἐφ’ ἡμῖν TA EF' hHMIN). We cannot control  
many of the events that govern our destiny, but we CAN control our  
attitudes toward those events. The Epicurean life-style is, to be  
sure, different from that of the Stoic, but there too is an emphasis  
upon emotional self-control; it is termed ATARAXIA (ἀταραξία)  
-- usually Englished as "imperturbability." It is freedom from  
distress (TARACH ταραχή).

Jesus teaches something a litte different but not altogether different:

Matt. 6:25 	Διὰ τοῦτο λέγω ὑμῖν· μὴ  
μεριμνᾶτε τῇ ψυχῇ ὑμῶν τί φάγητε ...
MH MERIMNATE THi YUCHi hUMWN TI FAGHTE ...

That active verb MERIMNAW is combined with another one, QORUBAZOMAI,  
which still is closer to TARATTOMAI in terms of mental distress (cf.  
Louw & Nida 25.234):

Luke 10:41 ... Μάρθα Μάρθα, μεριμνᾷς καὶ  
θορυβάζῃ περὶ πολλά, ...
MARQA, MARQA, MERIMNAiS KAI QORUBAZHi PERI POLLA ...


> The Friberg tags have all the MP forms of TARASSW as "passive", but  
> that carries no weight, because he also tags all the middle forms of  
> FOBEOMAI as passives.

I do think that the Friberg tags will be revised eventually (perhaps  
another 2 decades? I have been told that they'll drop the notion of  
deponents the next time ANLEX and AGNT are revised). We've been  
successful in getting Accordance to tag Greek verb forms in terms of  
their morphology rather than subjective semantic interpretation of the  
taggers.

Carl W. Conrad
Department of Classics, Washington University (Retired)






More information about the B-Greek mailing list