"WHAT IS CONSENSUS?" asks Michael Pahl. Here's my take. This is longer than I intended and I could probably keep going, but I do want to do some other things this evening. Feedback and criticisms welcome, as always. Drop me an e-mail or post on your blog in response.

First, consensus is a very slippery notion. What is consensus? I wouldn't think it's unanimity, at least not in dealing with such large and diverse groups of people. And it's certainly not simple majority. So what is it? 70%? 80%? Can 90+% safely be considered consensus?

I take consensus to mean that more people think so than not, but not everyone thinks so. Call it 51%-99%.

Second, who gets to be part of the polling sample? Only those working within a particular historical or theological perspective or methodological paradigm? All those who have published scholarly monographs on the subject? All scholars who have studied the subject in depth, whether they've published on it or not? And who determines what makes a "scholar" or appropriate authority on the subject?

I think most people measure consensus on the basis of published work in scholarly monographs and collections of essays, articles in peer-review journals, etc. I would weigh this more heavily than what a scholar thinks but hasn't published, since publication in itself requires more work and more commitment than just studying. When one refers in a publication to a consensus about something, normally a footnote follows with specific references to previous publications. Someone is a "'scholar' or appropriate authority" if he or she publishes monographs with publishers respected in the field or articles in relevant peer-review journals. Scholarly conversation has been going on for many, many years. "Scholars" and "authorities" are people who are familiar with this conversation; who have the necessary training (languages, knowledge of primary sources, methods of a field, etc.) for their opinions to be worth something; and who carry on the conversation in the venues that have developed to continue it. The process is far from perfect, but it works pretty well. And I don't know what one could replace it with which would work better.

Third, how does one actually go about doing the polling to assess consensus? One can meticulously search through all relevant literature on the subject and record opinions in a table (which is sometimes done on specific issues), but this then leads to some of the questions in the previous point, and one often quickly realizes that individual scholarly perspectives on any given issue are too complex to fit neatly into a simplistic table or scale.

So don't be simplistic. Read carefully and widely, particularly the published work of people who disagree with you, and be able to summarize the views of people you read with nuance and accuracy. It's hard work.

Finally, even if one can get past the previous questions, what does consensus prove? The consensus in 1976 on Paul's perspective on the Torah and his understanding of "justification by faith" was pretty strong, I would guess, until Ed Sanders provoked a Kuhnian paradigm shift with his Paul and Palestinian Judaism the following year. Now there is as much variegation in perspective on these issues as some of Sanders' respondents have claimed was evident in the nomism and soteriology of first century Judaism. Did the consensus prior to Sanders' book make that so-called "Lutheran" view correct? Does the lack of consensus since mean that no one has any real grasp of any of the issues?

The consensus is the thing you need to improve on if you want to advance the state of the question -- which is what scholarly research is all about. No one should pretend that a consensus is the final word. It's the starting point for your research. Apply new evidence or new methods or better reasoning to give us a better understanding of the problem. People who challenge Sanders's work don't want to go back to what was before it, they want to show that the problems are more complicated than he realized.

It's like hauling dirt up a hill so that you gradually make the hill higher so that everyone can see farther from its summit. (I like this analogy better than the one about standing on the shoulders of giants. Most people who came before us were not giants in comparison to the rest of us. But there were an awful lot of them and collectively they, each standing on the shoulders of those who came before, do constitute a kind of giant.)

Stipulated: a consensus can be an improvement on previous, wrong answers, and still turn out ultimately to be wrong itself. Newtonian physics was a vast improvement over anything that came before it and is still useful for virtually all everyday purposes today. But special and general relativity and quantum physics have shown that it is actually pretty much entirely wrong. Oh, and relativity theory and quantum physics both fit all the experimental data but are mutually incompatible, so one or both of them must be pretty much wrong too.

Stipulated: there are problems that are intractable with current evidence and on which we spin our wheels. Some of these are the questions we would most like to be able to answer. But there are plenty of problems on which we've made good progress too.

Some currently intractable problems, it seems to me, are the authorship of Joseph and Aseneth; the decipherment of Linear A; the origins of the Temple Scroll; and (to think big) the historical Jesus and the historical Josiah. Some will disagree with me about one or the other of these, but they won't be able to point to a consensus on what the answer is. Often (generally?) the problem in such cases is that we just don't have enough information to narrow down the possibilities very far.

Some random examples of problems on which we've made good, sometimes extraordinary, progress: the relative sequence and absolute dating of ancient ceramic typology in Palestine; the decipherment of ancient Egyptian, various languages in cuneiform (Akkadian, Sumerian, Hittite), Ugaritic, Phoenician, Nabatean, Palmyrene, etc.; the recovery of the previously almost entirely lost history of much of the ancient Near East from c. 2500 BCE to c. 500 BCE; and the basic decipherment of the Dead Sea Scrolls.

If there isn't a clear consensus, your mission, should you choose to accept it, is to find an answer that is more persuasive than the ones currently available, and that then becomes a consensus.

Until someone else improves on it.

UPDATE (14 July): Jim West replied to Michael Pahl here. (I really don't think things are as bad as all that.) Michael comments further with reference to my post here. (I don't see the difficulty with a "consensus on consensus." It's rare for people to agree unanimously on anything, but that doesn't mean we can't keep groping our way in the general direction of the truth. Let us steer between the Scylla of overconfidence in our own theories and the Charybdis of nihilism.) And Ed Cook and Mark Goodacre have thoughtful posts responding in detail to some of the issues that have been raised.

UPDATE (15 July): More here.