Source Criticism II

I found this lying about in my drafts, where it had lain idle for a goodly while, so I went through it, changed a few bits here and there, and here goes.  The approach is perhaps a bit circumlocutory, but please bear with me — it does become relevant :-)

For the past 9 years I have been working with mobile phones — particularly with their wireless performance (the ratio of erroneously received bits to the total number of transmitted bits) in a variety of situations. When I have analysed the performance of two phones, the inevitable question has always been ‘which is best?’ After I have carefully explained how one is better under some circumstances while the other is better in other circumstances, the next question has always been, ‘yes, but which one is best . . . overall?’

But how are we supposed to compare?

Should we compare the very best that one phone is capable of to the very best that another is capable of? Or perhaps we should compare the worst to the worst — for which phone does the largest fraction of test scenarios fail completely? Or should we perhaps compare the bulk of the data? In statistical terms the median^[1] or the mode^[2]? If we want a definitive answer, we can only compare one set of numbers.

I start with this illustration of a professional problem because it is, to my scientifically conditioned mind, very similar to the problem of saying something about the value of source criticism of Tolkien's work.

Based on the argumentation in Tolkien and the Study of His Sources edited by Jason Fisher, it is my impression that many would wish to dismiss source criticism as a legitimate critical approach because Tolkien himself disliked it, but I have never agreed with this argument: there are many points where I would disagree with Tolkien; in some cases I even appreciate Tolkien's views as an integral part of his sub-creation and despite disagreeing with him, I wouldn't have any adaptation that took a different position on it (this was my main complaint against the New Line Cinema adaptation of The Lord of the Rings). If new technology and new machines appear to make my life easier, I don't really care if it is at the cost of a few trees . . ..

There is, of course, the question of purpose: if we wish to do source studies, it is fair to ask why, and what we would wish to achieve by it. Verlyn Flieger, at the panel debate on source criticism at the Return of the Ring conference in Loughborough this August, suggested that the purpose was to understand the mind of the author (an endeavour I find no less daunting than Kristine Larsen's wish to “improve our individual chances of holding our own, if only for a brief moment, in a lively discussion with the Good Professor in whatever version of the Eagle and Child awaits the Second Born beyond the Walls of the World.”) In the panel, the conclusion was that we should do source criticism in order to better understand the “mental landscape” of the author. As landscapes can be understood in various levels of detail and abstraction (just play around for a bit with Google maps), I have no problem with that definition, and I would agree that it is a worthwhile effort: even if you wish to understand the story-internal origin of the Orkish race, you will need to understand Tolkien's mental landscape, and understanding it ever better can only help us in our pursuits to also achieve a better understanding of his sub-creational work.

The main issue that I have had with source criticism of Tolkien's work has been that so much of it has been so very badly conducted. I have, in my earlier post about source criticism, listed a number of the problems I have encountered in such studies and I will not expand on that here. One thing that Jason Fisher's book has done for me has been to open my eyes also to those few excellent to perfect examples of source studies that are also there while the bulk of the source studies are still poor. I've tried to illustrate this with a figure — the majority of source studies of Tolkien's work lie between the abominable and the tolerable (with most being merely poor), but a small fraction are good, excellent or even better. In the top we have a few studies such as Verlyn Flieger's Splintered Light.

The following paragraph is inserted after first publication in response to Jason's insightful comments below:
I should add that this represents my personal impression of “Tolkien Source Criticism” alone. As Jason Fisher points out in comments below, this is likely to also be true of other approaches, or indeed of all approaches (I do believe I could find other approaches for which I would say that the same picture to be true). This is also based on what is, after all, a limited sample (I have not read all Tolkien source studies) which may of course not be representative, and it is based on my personal ideas and preferences. I have targeted source studies specifically in response to reading Jason Fisher's book in which he sets out to improve on Tolkien source studies, not because I find that this approach shows particular problems compared to other approaches.

Looking at this, the big problem is upon what we should base our evaluation of source studies as a method for Tolkien criticism. Should we base it on the fact that a select few can use this approach for sublime results? I would venture that this is probably the result of the people rather than the method. Similarly I would claim that those conducting the abominable studies would probably do so regardless of the method. On the other hand, why should we judge the method on what the average guy can make of it?

When we realise that the quality of the resulting study will depend at least as much on the scholar and the specific topic as it does on the chosen method, why should we expect to be able to say anything general about the method at all? As in so many situations, the only answer is, ‘It depends!’

Jason Fisher's book, and in particular his own contribution to it, ‘Tolkien and Source Criticism: Remarking and Remaking’, gives a set of rules and sets up a standard that may help shift the distribution towards something less bottom-heavy — wouldn't it be wonderful if the bulk of Tolkien source-studies were good rather than poor? Certainly in Fisher's book itself the skew is the other way around and the bulk of the essays are better than tolerable.

While I shall probably still approach source studies with a certain degree of scepticism (so as to avoid disappointment), I will also do so with the small hope that this one might be one of the excellent studies, knowing that such studies do exist and are possibly not quite as rare as I had previously thought.

Oh! I do hope to have more to say about Fisher's book at a later point — this is not meant to be a review, but rather some further thoughts on my personal perception of source studies as a critical approach to Tolkien's work.


[1] The point where exactly half the results are better and half are worse. Somewhere between ‘Poor’ and ‘Tolerable’ in the figure.

[2] The point where the most results are concentrated — where the function peaks at ‘Poor’ on the figure.

Source Criticism

I am about to embark on reading Jason Fisher's new book, Tolkien and the Study of his Sources, and I thought that I would carry out a little experiment — I am, after all, an experimental physicist by education ;-)

Now, before I start reading, I will record some of my thoughts on source criticism, and when I have finished Jason's book on the practice, I'll post again to note any developments and changes in my position.

First off, what is source criticism? Having decided to do this experiment, I realize that I don't really know the formal definition as such. I know that it involves the identification of more or less probable sources of inspiration for elements in a story — e.g. Kullervo in the Kalevala as the original inspiration for Túrin Turambar — but I don't know if the term is used to cover more than just the identification of the possible source, so here at least there is something for me to learn ;-)  However, for the purpose of this post, I will use ‘source criticism’ to refer to the activity of identifying sources exclusively.

So, source criticism …

Generally my first criterion when evaluating literary criticism is whether it affects my appreciation of the work in question — I like it best if it can heighten my appreciation or deepen my understanding, but I guess that such a positive effect is a luxury one cannot always insist on :-)  Now, in my experience this is very rarely, if ever, achieved by source criticism alone (in the sense given above) — not that it doesn't occur in any study that takes its outset in source criticism, but the effect is then achieved by combining the source criticism with other, often comparative, approaches.

Tolkien discourages source criticism in e.g. his essay On Fairy-Stories, but that is, in my view, not a good reason in itself to abstain from it, even with regards to Tolkien's own writings — he also discourages the study of an author's biographical details, and that at least is a widely accepted and appreciated area also within Tolkien studies, including the study of biographical details as sources of inspiration for his fiction (where source study and biography meet), and I greatly appreciate good biographical studies that give me an increased understanding of the man behind the art.

This also implies one route in which source criticism can be expanded upon: if Tolkien (or some other author — my considerations are not limited to Tolkien, though his work is my focus) knew something and adapted (consciously or not) it into his own art in a given way, does this, then, tell us anything about the man himself? E.g. about his position with respect to the source?

Another route might be to turn the focus to the source itself. The mere fact that Tolkien may have known some other work doesn't necessarily mean that it is interesting to me, but some further description and criticism of the source might make me appreciate the source more (or make me interested in experiencing the source personally).

One can also imagine the route of speculative extrapolation: if indeed Tolkien based some element of his fiction on a certain source, can we then use that assumption to tell us something about this element in Tolkien's fiction, that is not obvious from a study only of Tolkien's work itself?

It should not be too difficult to expand upon this list of routes, but the common characteristic is that even if they do take their outset in source criticism, they all move beyond the mere identification of a possible source and add something that is not, in the sense I use it here, source criticism. When used in such ways, I am all in favour of source criticism as a good and sound basis for critical studies, and I have thoroughly enjoyed many such studies (e.g. by Verlyn Flieger and Tom Shippey to name a couple of the most prominent scholars in the field).


Unfortunately that is not always the case.

Much of the source criticism that I have seen has been self-congratulatorily satisfied with identifying a source, demonstrating the erudition of the critic rather than attempting to understand Tolkien's work in any greater depth. These are, I realize, quite harsh words, but I must insist that the mere identification of a possible source and noting a number of more or less obvious (and occasionally strained) similarities is, in and of itself, quite uninteresting — it is only when the study is expanded beyond the mere source identification that it has the power to add value for the reader (exceptions can probably be found to this blanket rejection, but I do believe it is true for the vast majority of possible source critical studies).

This also brings me to speak of the problematic elements that mar at least some source critical studies. I will refrain from citing explicit examples of the various types of problematic behaviour, and so you will have to trust me that none of this is my own invention.

My personal background is in the sciences: I hold a master's degree in physics and a bachelor in computer science, and I work with the mathematical and statistical analysis of industrial tests and test results. As such there is often one particular thing that bothers me in source studies: when they ignore to consider alternative hypotheses (including the possibility of an amalgam of different sources, all pointing in roughly the same direction and alloyed in the crucible of the author's creative imagination). It is fine that they identify some source that may have inspired Tolkien, but if they don't consider alternative hypotheses at all, they can only show that a connection is possible, but say nothing about its likelihood: it is actually possible that I will suddenly tunnel a metre into the air, but it is not at all likely. One alternative that should always be considered is that of noise — that any similarities are merely random (I haven't come across a study that tries to quantify the amount of similarities between randomly selected works, but that could surely be interesting).

This is probably tied closely to the idea seems to underlie some source critical studies, namely that everything must have a source. ‘What is the source of this?’ the critic asks rhetorically, implying that they are not willing to consider the idea that there is no source — that we may be dealing with original invention. If we were dealing with original invention, then the similarities would be random, which is why it would be good to know something about this ‘noise-level’ in literature. I should probably add that I do not think that this is intentional — I don't think that any serious source-hunter would deny Tolkien's originality and inventive imagination, but their focus on the sources makes it easy to accidentally alienate some readers by appearing to do so.

A last practice is definitely not unique to source studies, and may not even be more prevalent there than in other critical approaches, but I will include it here because I have seen it in source studies, and I believe it is an example of mild misconduct in scholarship. I am talking of the practice of stating unconnected (and incontrovertible) facts in such a way that the reader is invited to draw the conclusion that they are connected, even though the scholar cannot make this conclusion her- or himself because there is actually no evidence for such a conclusion. The form it takes in source studies is usually to state facts showing that some source (e.g. in the form of a book) was available to Tolkien, for instance by noting that a book was available in a library at a period when Tolkien had access to, or even was using, said library. This invites the reader to conclude that Tolkien not only had access to the book, but also did access it: a conclusion that is completely without basis — in particular since any evidence that Tolkien did access it (e.g. that he took it out from the library) would surely have been given if it existed.


So, this is where I start. Now I look very much forward to reading Jason Fisher's book — I suspect that the definition that I have used of ‘source criticism’ is too narrow, but I have employed it here as a useful way to speak of the source-finding activity alone. It may seem odd, but I hope to be proven wrong, to be able to see the usefulness of source criticism (i.e. the source-hunting alone, as I have defined it above): it is much as with cheese, which I, unfortunately, intensely dislike, but watching friends and family enjoy a good cheese table (with a good red wine, which of course I can share the enjoyment of), I can get the feeling that my tastes prevent me from enjoying something very valuable. I trust, however, that my reason is more easily persuaded by reason than my taste buds ;-)