I appreciate Mark Goodacre's response to my recent posts. But I must insist that experimentation makes a difference - a huge difference - in what we can and can't know, and what we can and can't argue. On this see the good post by N.T. Wrong (who seems to be consistently right - when will we know your ID?).
Take for example Professor Bartlett's experiments which were created not as laboratory experiments, but as real life tests of human memory. He asked subjects to read a story three times. Then at varying intervals he asked his subjects to write the story down. He noted specific alterations (what we call today memory distortions) to the materials. These alterations are specific and consistent. I found the same ones, and I did not have access to his experiments until after I had run my own and was writing up my analyses. This was another shock to my system.
What is fascinating is that originally I wanted to track the differences in media environments. So I had divided my subjects into five groups: oral-to-oral (OO=heard saying and orated reproduction); oral-to-written (OW=heard saying and wrote down reproduction); written-to-oral (WO=read saying and orated reproduction); written-to-written (WW=read saying and wrote down reproduction); and sources-retained (SR=read saying and wrote down reproduction but retained written "original"). In other words, I had four media environments in which composition and recall relied entirely on human memory, while one media environment allowed my subjects to consult a written source. I was confident that I would discover all sorts of differences in the first four mixed media reproductions and was ready to track them.
Much to my surprise, the differences in these four media environments were not present at all in long-term memory reproductions. In other words, in the four the media modes that relied on human memory for transmission, there was no difference in how the material looked after it was transmitted. The material underwent the same type of changes at the same rates. The differences occurred only when the written source was retained and the subjects could consult it.
In other words, from the reproductions themselves made by my subjects it was impossible to deduce whether the subject heard the saying or read the saying, wrote it down or orated it, as long as this was done from memory and not from consultation of a written source. The factor for distortion in transmission was NOT the media environment - it was dependence on human memory.
So how can we tell if the author of Thomas' parable of the Wheat and Tares was dependent on Matthew? If we are dealing with literary dependence through consultation of Matthew's text, we would expect either near verbatim reproduction or paraphrase according to the results of my experimental exercises - and we have neither of these. This is too complicated for me to go into here, so please have a read of my article and the results which I charted.
If the author of Thomas' parable was remembering Matthew's version, then the only way to know that it is Matthew's version and not some other is to detect a significant amount of Matthean development of the parable in Thomas' version. It becomes difficult at this point to detect exactly what is Matthew's development, since the concept of ONE originating parable cannot exist in an oral-rhetorical culture.
I REALLY caution all of us on this point, because I discovered in another experimental exercise that I didn't publish (yet?) that when you have twenty-five versions of a parable in front of you that all look similar, if you ask how these twenty-five versions came about, you are tempted to try to build a family-tree based on similarities in some versions. But what I'm finding is that people make the same adjustments to versions INDEPENDENTLY of each other, and these adjustments are due to the way in which our memories work, and cultural and linguistic phenomena within a given generation of people. So the twenty-five versions may be all independent versions with no relationship to each other except that all persons were present to hear a version orated (and in fact were in the case of my experimental exercise). I admit being quite shocked about this, and seeing form criticism fail miserably before my very eyes.
But if we examine the Thomasine parable carefully we see that we do not have significant commonalities with Matthean secondary developments (="rationalization" or "idiosyncrisies" as Bartlett would have phrased it). The verses that appear to me to reflect Matthew's theological interests - the dialogue of the enemy in vvs. 27-28 and the accumulated proverb in v. 30 - are not found in Thomas' version. But Thomas' version represents a condensed form of the parable, although it is impossible to conclude that this originating form is a memory of Matthew's version or some other version available to the person who composed Thomas' version (which still, based on studies of oral composition, has the characteristics of an orally-composed text).
6 comments:
I still wonder if doing the same tests in non literate countries would produce different results or not.
I also wonder in literacy and non literacy are hard wired in the sense that if someone has never learned to read puts the information in different parts of the brain and if that affects memory.
I know that body to brain and brain to body information affects not only information stored but also brain chemistry. So do events in life. Depression is affected by life experience as is life affected by depression.
And rerouting of information because of trauma changes brain patterns too.
In other words I think sampling those only who can read may produce skewed results. Those who rely on hearing memory instead of sight and reading may use different brain tracks and may retain memories better, worse, or in a different way than those who rely on sight and reading.
I think if may be a bigger subject because of current brain research than we have any idea.
And yes, as you have pointed out, Dr. Deconick, it affects form criticism and possibly other areas as well.
Bob,
Keep in mind that we are not comparing literates with illiterates when we attempt to generalize to the early Christian world on the basis of modern experiments. We can not study orality of early Christians, because we are not there to record what they are saying or hearing. All we can do is study their texts. Matthew was literate, Thomas was literate.
Still, it remains very difficult to generalize to early Christians, because what they are attempting to do is so incredibly different from what we require from our modern subjects.
Also, compositional procedures were very different and so was the memory framework that affected the encoding and retrieval from memory. That is, how a person processes a new text depends to a large extent on what is already in memory. For example, if as person already has a very good memory of septuagintal texts this will affect how the person processes a new story. It may remind the person of the older texts (memory is associative) and this could have several different kinds of effects on the memorization and on the composition of a new text.
I think one thing that it would be fascinating to study, but it might be extremely difficult to arrange an experiment in our society, is the extent of similarity between the following two scenarios: (1) a story is read repeatedly, word for word, from a Gospel or other such source, over the course of several years, and then another person is asked to retell it or write it down; (2) a story is repeatedly told without there being a written version with set wording, and then another person is asked to retell it or write it down.
I wonder whether the sort of repetition that is possible only over long periods of time, and the precise repeated wording possible only when one has an authoritative text that is read aloud to a congregation regularly, make a difference in the long run. I suspect that the results might well be the same as in your experiments (which I assume were shorter-term, but I may be wrong about that), but I also think that there might be key words or phrases that might get stuck in the mind through repetition, that would not in purely oral transmission because that phraseology might never be used a second time.
As I write this, I am starting to wonder whether the conclusion that we need to draw is that there are so many different possible modes of interconnectedness between texts and their oral environment, that we cannot in most cases figure out precisely the relationship. And that uncertainty might be more revolutionary for NT studies than any firmer conclusion could have been!
It was good to hear the preliminary results of this study, which prompt a few questions and comments. In the SR group, was the participant supposed to not consult the source except at spots where memory failed? I.e., it was not quite the same as transcribing, I presume?
Another question: Were any of the participants not already well aware of the Mustard-seed parable in Mark (or Matthew) that was involved, before you initiated the test, as opposed to the others? If so, one would need to separate the results into those two groups at least before further analysis.
A complication: By my understanding, a similar study needs to be contemplated in which the written source, in WW, is in a different language than the write-down language (i.e. W’W), as translation adds much variability. Since the external evidence is of Matthew being first and in the Hebrew tongue with the other texts being written in Greek and based on this Matthew, and since the Aramaisms and Hebraisms within Matthew cannot be denied and would have carried over into Mark and Luke also, and since Mark could even more easily have been an abbreviation rather than Matthew an expansion, this needs consideration. Parallel texts already exist, however, for which W’W can be studied, leading me to conclude that the translator of Hebraic Matthew into Greek utilized Mark and Luke, and purposely replicated lengthy strings of words.
Another factor in memory: it is my experience that no matter how many times people hear the Lukan Christmas narrative they are convinced that Mary rode a donkey, that Joseph and Mary spoke with the Innkeeper, that the manger was in a shed of some sort and that there were animals at the scene of birth.
Memory of the text has been corrupted by other sources of information like Sunday School classes and Christmas pageants.
Which, of course might suggest a similar corruption (although in different ways) in the early Church's memory.
On other comment: I think one needs to work with people who do not come out of a tradition that uses Scripture memory. These folks might have perfect memory of an assigned passage, although I suspect it would be in the King James Version.
I had to memorize Scripture when younger and my memorized passages are all in the Revised Standard except for a few in the King James. It drives me crazy to read a passage in worship in the New International Versionthat I have memorized because it just doesn't sound right.
Post a Comment