What makes corporate blogs (cross-culturally) successful?

2009 February 19
by Cornelius

I’m still working on a lengthy post in which I want to summarize the most important findings of my research on corporate blogs, but in the meantime here’s another perspective on the topic from Nils König. Take the time to have a look at Nils’ paper – the section on f-scores was naturally intriguing to me.

I am hesitant (as is Nils, if I interpret his comments correctly) to assume that the f-score differences between English/German and Russian are really anything other than systemically conditioned. The difference in score is more likely to be a difference between the three languages than a difference between culturally unique blogging practices.

It is also notable that Nils’ f-score results are quite high – 68.6 for English-language corporate blogs compared to only 53.3 in Nowson et al’s original investigation of the f-measure in (non-corporate) blogs. My own scores are somewhere in between, though with a lot of variation. Whatever the cause, 68.6 is unusually high, outranking even some genres of academic writing according to Scott and the BNC.

Both this and the high scores for Russian point to the f-score’s limitations more than to any other result in my view. The measure reflects word class distributions and derives predictions on the qualities of texts (“formal” vs “contextual”) from them. But language is compositional, word classes are an indealization of a language’s lexico-grammar and “formal” in Heylighen and Dewaele’s definition does not mean “stiff” or “dry” – it means explicit and context-independant vs. implicit and context-dependent.

In other words, I’ve become quite careful with assigning too much argumentative force to the numbers as such. Nils’ observations on the other hand sound interesting and very plausible to my ears. Personally, I’ve been moving more into the qualitative direction lately when it comes to methodology, simply because measures like the f-score are so fickle. This paper on intercultural pragmatics that we discussed in my class on the linguistics of business communication last semester offers some very interesting observations on how cultural influences shape genres using examples.

Close reading, while tedious, can be superior to a corpus-linguistic approach sometimes. On the other hand, Nils’ gave me an idea for another context in which the f-score might prove to be a very useful tool. Thanks for that :-)

from → Blogging Studies, Corporate Blogging, Style

1 Comment
2009 February 20
Nils König permalink

Hey Cornelius,
The use of the F-score was of course inspired by some of your writings in the first place (remember our e-mail discussions about 8 or 9 months ago?). I guess the scores vary depending on the tagsets used, so I refrained from comparing them to existing studies (as they vary greatly as well) and only limited it to a relative evaluation. Regarding the analysis, my conclusion was actually that the differences are NOT comparable among cultures, because they are very likely affected by language characteristics (though there is no empirical data on that). The only valid conclusion is that blogs are indeed less formal than press releases (no surprise about that).
I absolutely agree that the f-score is very limited in its application and qualitative aspects are more likely to be important in determining blog performance. Hence I included a range of other variables, with some going in the qualitative direction. Currently, I’m working on a two further approaches to that issue. Still, the trip into linguistics was quite interesting :-)

Comments are closed for this entry.