2020-11-25

Did Jonathan Z. Smith Really Not Understand Ideal Types? (Part 1)

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.

by Tim Widowfield

Religious historian Jonathan Z. Smith

Over the summer and autumn of 2020, I’ve been catching up and rereading several important books on the New Testament, especially those that have approached their subjects from a sociological standpoint. Those works led me to others (sometimes the bibliography is more worthwhile than the book itself), and so on.

I remember reading Jonathan Z. Smith and noticing what he had actually written did not correspond well with what Robert M. Price had told us he wrote. Price has continued to insist for many years that Smith didn’t understand Weberian ideal types and that if an instance of a type did not conform exactly to the type, then we had to discard the instance.

Yet, in Drudgery Divine we observe in Smith’s writing an honest effort to categorize unique events within frameworks of classification. In fact, he pushed against “uniqueness” as a modern concept, too often used as an excuse to mystify, a lazy justification not to compare, for example, one event with another.

Let us be clear at the outset. There is a quite ordinary sense in which the term ‘unique’ may be applied in disciplinary contexts. When the historian speaks of unique events, the taxonomist of the unique differentium that allows the classification of this or that plant or animal species, the geographer of the unique physiognomy of a particular place, or the linguist of each human utterance as unique, he or she is asserting a reciprocal notion which confers no special status, nor does it deny–indeed, it demands–enterprises of classification and interpretation. A is unique with respect to B, in this sense, requires the assertion that B is, likewise, unique with respect to A, and so forth. In such formulations ‘uniqueness’ is generic and commonplace rather than being some odd point of pride. In my language, I would prefer, in such instances, the term ‘individual’, which permits the affirmation of difference while insisting on the notion of belonging to a class. [pp. 36-37, emphasis mine]

He tackled the subject of categorization and classification in greater depth in his 1982 work, Imagining Religion. When trying to explain what a religion is and how one particular religion fits within a framework of categorization, we often stumble on the problem of necessary and sufficient criteria. He wrote:

Perhaps the most promising development in recent years has been the return to the suggestion made by Michel Adanson in 1763 in controversy with Linnaeus over the question of the mutability of species. Adanson argued that the members of a given taxon need not possess all of the defining characteristics of that taxon and that there was no a priori justification for deciding what characteristics were most definitive. Although the state of the art in the eighteenth century was not such that Adanson could have put it this way, what he was suggesting was something like a statistical approach to classification. [p. 4, emphasis mine]

Here we see the way out of the quandary in which it seems as if all religions ought to have gods of some sort at their cores, and yet — Buddhism, and yet — Jainism.

Smith referred to this system of categorization as a “polythetic mode of classification.” By this means, we can leave aside the unhelpful procrustean bed of rigid classification. In his words, we could “surrender . . . the idea of perfect, unique, single differentia.” And further, with large numbers of individuals, we would expect to find (as in evolutionary biology) examples in which one individual might share hardly any characteristics with another individual. In fact, they might not share any at all.

To state this more concretely. Imagine a group of six individuals, each possessing three characteristics of a set, A-H. Individual 1 has characteristics A, B, C; individual 2 has B, C, D; individual 3 has A, B, D; individual 4 has A, C, D; individuals 5 and 6 have characteristics F, G, H in common. Individuals 1-4 would be formed into a polythetic group sharing a number of characteristics, although no one characteristic is found in all four individuals. Hence, no one characteristic is definitive. Individuals 5 and 6 form a classic monothetic set with the only question remaining the determination as to whether characteristic F or G or H is definitive. [pp. 4-5]

To put it another way, various individual examples of a type may not share a single prominent characteristic, and yet they may still firmly sit within the general boundaries of the ideal type.

With all this in mind, we may begin to doubt Price’s accusations against Jonathan Smith. He has stated them several times, most recently in his 2018 book, Bart Ehrman Interpreted:

[Jonathan Z.] Smith accuses previous scholars of cobbling together disparate features of very different myths in order to construct an artificial synthesis, a chimera which does not resemble any actual myths or gods. He is shocked, shocked, to find significant differences between the so-called dying-and-rising god myths and declares the patient dead with no resurrection on the schedule. In other words, if the various myths of Osiris, Attis, Adonis, et. [sic] al., do not all conform to type exactly, then they do not fit into the same box, so let’s throw out the box. Without everything in common, Smith sees nothing in common. [p. 194]

How can we square this “bad” J.Z. Smith who throws babies out with the bath with the “good” Jonathan Smith above who argued for a polythetic classification system to help us understand complex realities? In the next few posts, I’ll take a closer look at what Smith really thought about dying-and-rising gods, ideal types, and the general process of categorization and comparison.


Smith, Jonathan Z. Drudgery Divine: On the Comparison of Early Christianities and the Religions of Late Antiquity. London: School of Oriental and African Studies, Univ. of London, 1990.

Smith, Jonathan Z. Imagining Religion: From Babylon to Jonestown. Chicago: University of Chicago Press, 1982.

Price, Robert M. Bart Ehrman Interpreted: How One Radical New Testament Scholar Understands Another. Durham, North Carolina, Pitchstone Publishing, 2018.

The following two tabs change content below.

Tim Widowfield

Tim is a retired vagabond who lives with his wife and multiple cats in a 20-year-old motor home. To read more about Tim, see our About page.

Latest posts by Tim Widowfield (see all)



If you enjoyed this post, please consider donating to Vridar. Thanks!


9 thoughts on “Did Jonathan Z. Smith Really Not Understand Ideal Types? (Part 1)”

  1. This may be another “strange Jekyll and Hyde situation”.

    • Godfrey, Neil (24 September 2019). “Review part 3: Questioning the Historicity of Jesus / Lataster (Addressing the Case FOR)”. Vridar.

    Lataster dissects each of the above failings in Did Jesus Exist? but interestingly goes further and contrasts Ehrman’s failings there with his books written before and after that one:

    Before and after writing that book, Ehrman was and is capable of proper critical research on the biblical texts. But for some reason, during the writing of Did Jesus Exist?, Ehrman’s standards dropped remarkably, only for the ‘old Ehrman’ to return soon after, as if he suffered from a fugue state. I suspect that Ehrman consciously or unconsciously realised that the case for Jesus would be very poor indeed if he consistently applied his critical approach and all of his vast knowledge to this question, leading to this strange Jekyll and Hyde situation. (p. 71)

    1. If Carrier is correct that per the Dying-and-Rising God Mytheme, Jonathan Z. Smith “didn’t even address 99% of the evidence for it, but flat out ignored almost all of it”.

      Then I want to know if Smith had an on the record position, in clear and unambiguous language, without equivocation for the Historicity of the resurrection of Jesus?

      • Carrier (19 September 2016). “Virgin Birth: It’s Pagan, Guys. Get Over It”. Richard Carrier Blogs.

      [Per the Dying-and-Rising God Mytheme] Jonathan Z. Smith did not refute that mytheme; he didn’t even address 99% of the evidence for it, but flat out ignored almost all of it, and focused on only one obscure and consequently irrelevant example—much as did, also, N.T. Wright

  2. As a side note: I concur with R. M. Price that the dying and rising god thing does not significantly reflect on mythicism v. historicism.

    “Robert M. Price & Christopher Hansen – INTERVIEW!”. YouTube. MythVision Podcast. 2019.

    [01:46:30] Ultimately I don’t think the dying and rising god thing, though fascinating, really bears on mythicism, because Rudolf Bultmann and Joseph McCabe and various others have long said there were dying and rising god myths and they were among the resources early Christians used to mythologize the historical Jesus.

    Bultmann goes into all of this stuff but he thinks, there was a historical Jesus, it was just he was made over in this image as he was: with a gnostic redeemer and the Jewish Messiah. If you could prove that there were dependencies of genealogical relationship, that wouldn’t really reflect on mythicism verses historicism anyway. So in a way it’s like a moot point—as fascinating as it is.

    Cf. Godfrey, Neil (29 May 2019). “Robert Price and Christopher Hansen Discussion”. Vridar.

  3. As a side note: Cook, contra Smith, asserts that “the continued use of the category of dying and rising gods” is justified.

    • Cook, John Granger (2018). Empty tomb, resurrection, apotheosis. Tübingen. ISBN 978-3-16-156503-8.

    [§.Conclusion] The review in this chapter thoroughly justifies the continued use of the category of dying and rising gods. [468] The resurrection of Osiris is the closest analogy to the resurrection of Jesus, although Osiris remains in the netherworld—wherever it is located. Horus’s resurrection is a clear analogy. The rebirth or resurrection of Dionysus also provides a fairly close analogy to the resurrection of Jesus. The revival of Heracles and probably that of Melqart are also strong analogies. —(p. 143)

    [468] pace D. Frankfurter, rev. of Mettinger, Riddle, Bryn Mawr Classical Review 2002.09.07 who argues that the category “dying/rising god” is merely a “Christian theological holdover, like ‘sacrament’ or ‘faith,’ from a time when all comparison was meant to legitimize or delegitimize dogma.” That premise is based on an argumentum ad hominem (i.e., those intending to “legitimize/delegitimize” dogma) and so is logically fallacious. The language is unavoidable, and even J. Z. Smith succumbs (cf. the quote in § 6.1 above from Drudgery, 133): there are “second to fourth century AD reinterpretation, within some of the ‘mystery’ cults, of archaic locative traditions of dead deities in new experimental modes which appear to testify to these deities returning to life. In the case of Attis, there are only scattered hints of this process.”

    1. “Jonathan Z. Smith, Drudgery Divine. depts.drew.edu. “Reviewed by Robert M. Price. Institute for Higher Critical Studies JHC 3/1 (Spring, 1996), 137-145.”

      Smith declares the famous “dying and rising god” mytheme a modern myth . . . not an ancient one.
      […]
      Smith, in Drudgery Divine, does come down, it seems to me, as an advocate of the principle of analogy.

Leave a Comment

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.