Homer and the Value of Statistical Analysis

Image

Homer by Rembrandt [Public Domain], via Wikimedia Commons

Several weeks ago, a number of news sites reported on a study undertaken at the University of Reading, which examined the dating of Homer’s Iliad.  Some of the headlines included:

“Homeric Epics Were Written in 762 BCE, Give or Take, New Study Suggests”

“Homer’s great literary masterpieces dated by study of Greek language evolution”

 “Geneticists Estimate Publication Date Of The ‘Iliad’”

Upon seeing these headlines, a couple of thoughts popped into my head. First of all, it was rather surprising this study had apparently been able to assign such a definite date to a piece of literature that had previously caused so many headaches among scholars. Nevertheless, with lines such as “Homer’s great masterpieces, The Iliad and The Odyssey, have been dated to around 762 BCE by new research…” it seemed as though a major breakthrough had been achieved.

Closer inspection of the study itself, however, revealed this to be an over-eager misrepresentation of the results (a link to the published article discussing the results can be found at the bottom of this piece). In an industry that requires attention-grabbing headlines in order to attract readership and drive website traffic, it comes as little surprise when media eschew more appropriate headlines; by comparison, “Homer’s Iliad may have been written at some point between 1157 and 376 BCE” is a rather lacklustre lead. Examples of similarly-definite (although ultimately misleading) headlines are common, with the claim from 2011 that faster-than-light particles had been discovered being a particularly troubling case (it turned out the scientists had made a mistake).

The second thought has stuck with me a bit longer, namely: why is this being reported at all? Leaving aside questions regarding whether ‘Homer’ was a historical figure and ‘his’ personal dating, the general consensus among Classicists and literary scholars has been for many years that the Iliad was first written down around the 8th century BCE. As in any scholarly debate, there are those who disagree with this view, but it is what is being taught to undergraduates – and what I learned in my intro to Classical literature course oh-so-long ago – and it can be considered a generally-accepted fact at this point.

Why, then, is a study that only supports existing concepts being reported so widely? I doubt, for instance, if the causes of polio were re-examined, and the conclusions were essentially “Yep, it’s exactly what we thought a hundred years ago”, that such a study would receive any attention whatsoever. It is certainly helpful to have supporting evidence, but it isn’t a ground-breaking revelation requiring international attention.

Accordingly, the reason for all of the attention surrounding the Reading study is almost certainly related to their methodology, as opposed to the actual results; statistical analysis is a ‘hard’ discipline – in contrast with the ‘softer’ Classics – and therefore the results of such an analysis may be viewed as more scientific and more reliable. But this is a profoundly problematic attitude as it ignores (or at least significantly diminishes) the efforts of countless scholars who have studied Homer and the Iliad for their entire professional careers. These researchers have often operated within the humanities, but that does not mean their results should be viewed with any less validity than their scientific counterparts. The humanities are a rigorous discipline as well, with similar long periods of study, emphasis on peer-reviewed work, and attention to detail. We just don’t work in labs…usually.

In other words, ‘scientific’ does not automatically equal definitive, and the work of Classical scholars should not be brushed aside so quickly. There are professionals in the humanities too!

The full study can be found here: Altschuer, E. L., Calude, A. S., Meade, A. and Pagel, M. (2013), “Linguistic evidence supports date for Homeric epic.” BioEssays. DOI: 10.1002/bies.201200165

It should be accessible to all, considering I am in middle-of-nowhere Canada and I can access it.

Advertisements

11 responses to “Homer and the Value of Statistical Analysis

  1. > “the general consensus among Classicists and literary scholars has been for many years that the Iliad was first written down around the 8th century BCE … it can be considered a generally-accepted fact at this point.”

    So, because it’s mostly agreed upon and considered accurate, it should be considered fact? This study presents a manner of investigating the question which is not subject to “truth by consensus.”

    > “Why, then, is a study that only supports existing concepts being reported so widely?”

    Because it, as the authors indicate, presents a powerful method of dating literary texts which relies strictly on linguistic evidence. As long as the information on cognates is accurate (i.e. the data are reliable, something that all sciences rely on), the method is valid, and the results are not subject to interpretive bias. We can estimate probably times, give ranges that have high probability, identify estimates which are very unlikely, etc. They don’t even need historical, cultural, or archeological evidence, strictly linguistics is enough to derive an estimate and a probable range.

    Nowhere does the article suggest that the Classics are less valid, or that the discipline is any less rigorous than the sciences. Several times it mentions that their estimates align with those of the Classicists. They’re simply acknowledging – as you have – that analysis in the Classics has a degree of subjectivity, and they are presenting an objective method which can help to date or confirm the datings of ancient texts.

    • I must admit, I wrote up a full response to you Casey, but it was far too long – and far too meandering! It has, however, given me plenty to think about and I may just write more about it at a later date. For now, however, I will try to be concise.

      The ‘general consensus = fact’ description was a simplistic way of presenting the issue. In reality, a more accurate comparison may be made with something like gravity. Namely, there are many things we know about Homer, the Iliad and gravity, but there are also many big holes in our understanding and different people explain these holes in different ways. Moreover, there are some explanations with more weight than others, but this certainly doesn’t represent true ‘fact’.

      Furthermore, I am not discounting the results of the study – or indeed the methodology of the study either. I think it is a valuable contribution to Homeric studies, and, as you mention, if the data are reliable then the study could present an interesting new approach to a very old question.

      My problem, however, is precisely with the attitude that ‘scientific’ studies such as those undertaken at Reading are a perfectly “objective method” – as you have put it – and the resulting conclusions are ‘better’ than could otherwise be achieved. I am quick to add that I doubt this is the attitude of the researchers themselves, and my issue is primarily with the media and societal views; the Reading study is merely the latest example. Far too often years (sometimes decades) of valuable and research minimized or ignored simply because they are a product of a humanities-based study.

      Well, so much for concision.

  2. Also, what’s the point of doing any sort of study if the degree of accuracy will be: it was year X, give or take 400 years?

    By the way, thanks for liking one of my posts!

    • Well, I’m not going to presume to understand the statistical models used by the Reading team without seriously brushing up on my maths, so my issue is not so much with the results. But, yes, there’s rather quite a lot of stuff happening in an 800 year period!

      And my pleasure, it was a great read! Thank you for coming to comment on my own.

  3. I agree with your first comment on the articles. This is certainly an example of journalistic sensationalism in a field that has increasingly become over-saturated with attention-grabbing headlines. If any argument at all were to be made, a title such as “New Method of Dating Historical Texts Developed” would be considered more relevant than “Homeric Epics Were Written in 762 BCE, Give or Take, New Study Suggests”. Overall I agree when you say that this is not information that needs to be broadcasted to the general public. That being said I also see these articles as an attempt by the media to report more ‘intelligent’ news than, say: http://www.huffingtonpost.com/2013/04/05/charlie-sheen-denise-rich_n_3022487.html (however I’m not ready to give out any ribbons just yet).
    On the subject of Humanities vs. Science (despite my obvious bias): I would still make the argument that you are extrapolating. For one, ” ‘scientific’ does not automatically equal definitive”. Well, yes. I don’t think any self-regarding scientist (or journalist I hope!) would ever make this claim. That being said scientific methods produce much more accurate – and, more importantly, more precise – results. This doesn’t make scientific research more respectable than that of a classicist, nor do I think that this is the opinion of any of these authors. It means only that scientific results – by some paradox of journalism – can more easily be sensationalized and misrepresented.
    In reference to the method of dating you suggest “such an analysis may be viewed as more scientific and more reliable”. And I would agree because it is, in fact, more reliable. But again I would argue that these articles should be attributed to the inherent nature of journalism rather than a widespread bias against the humanities in society.
    Also, science is just cooler. Classics doesn’t really stand a chance when we have things like the ‘god’ particle and the big bang (both of which, ironically, was popularly named after unwittingly being misrepresented in the media).
    Just some ideas, let me know what you think!

    • Thanks for the comment, Connor! There are some good thoughts in there, but I have a few issues.

      “And I would agree because it is, in fact, more reliable.”

      A scientific approach is undoubtedly preferable in many instances when dealing with certain questions, including investigations of the Higgs boson and the Big Bang. I don’t, however, think it is the best – and certainly not the definitive – approach on every occasion. So what makes it more reliable on this occasion, and in other similar cases, than decades of research by Classical scholars?

      Moreover, I dislike the suggestion of a “Humanities vs. Science” dichotomy; the two are not, and should not be, exclusive. In modern research, including the Reading study, the two often go hand-in-hand to the ultimate betterment of the results. I am not trying to elevate Classics or the Humanities above hard science, merely pointing out the pervasive attitude that ‘science’ is more reliable – which you yourself have so kindly demonstrated.

      • Okay, that’s fair. Certainly we shouldn’t think of the two fields as exclusive like you suggest. And I think this debate on whether science is more ‘reliable’ stems simply different uses of the word itself. For, almost by definition, science produces the most ‘reliable’ results i.e the results with the smallest uncertainty. For example perhaps they were able to date the Iliad to within 100 years as opposed to within 500 years previously.
        But as you say, this scientific method is not always the most appropriate. Sure we may be able to date historical scripts, but if, instead, you’re want to study how the Iliad influenced major philosophical thinkers of that era…. well, that’s best left to a classicist – for they would produce more ‘reliable’ results.

  4. For some reason, I can’t reply directly to your latest comment, Connor, so I’ll just write it here instead.

    You still haven’t demonstrated what makes the Reading study more reliable than the pre-existing research; both methods reached the same conclusion.

  5. When doing a research or study, precise answer must be the one given. People may doubt something about the study whether it’s true or not. Exact details or facts is what people need. What is the point of studying/researching the study if they cannot give an accurate answer?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s