Why comparing alternative transcripts doesn’t necessarily yield the truth of what was said

You may have seen some rather hilarious ‘alternative lyrics’ for Carl Orff’s famous O Fortuna that are circulating on the internet. As with many forms of word play, these are not only entertaining but also give some important insights regarding language and speech – and, in this case, into the effect of priming on evaluation of forensic transcripts.

The thing is, there are by now quite a number of different parodies suggesting a selection of quite different lyrics. If you like this kind of thing, you can get dozens by searching YouTube for ‘o fortuna misheard lyrics‘.

But we have embedded a few below for your convenience. You can study them to determine

  • which are funniest
  • which fit the acoustics of the words better.

But they also give some insights that are highly relevant to forensic transcription.

The funny videos

OK – so with all that build-up, here are the videos. Watch as much as you can stand, then scroll down for their relevance to forensic transcription.

Parody 1

Parody 2

Parody 3

A more important observation

Some people have suggested that to obtain a fair evaluation of poor quality covert recordings, we should not look at the police transcript. Rather we should obtain a range of transcripts, then perform acoustic analysis and choose the transcript with the highest statistical likelihood of being correct. Would that ensure we prevented inaccurate transcripts from influencing the jury?

Imagine you were given O Fortuna as evidence in a case (we leave you to dream up what the crime might have been) and asked to evaluate the various youtube transcripts for reliability.

Would choosing the ‘most likely to be correct’ of the alternatives you have just heard get you close to the real lyrics?

For those who would like to study this in a bit more detail

In case you would like to listen without the distraction of the videos, here is the music alone (just the first 30secs).

In case you are not a Carl Orff scholar, here is what appears to be the correct (Latin) lyrics with English translation.

But what you might find even more interesting is to play the music and let your eyes follow alternative transcripts.

To facilitate that, we have put together several texts side by side (below). Once you have the music going, you can click the text to open the page in a new window for easy reading.

You may notice how (if you are like most people) focusing on one transcript rather than another can really make it seem that the words fit the acoustics.

It is fun to observe the way your perception is ‘primed’ as you switch between different transcripts, especially when you know the transcripts are humorous parodies.

Much harder to notice the priming effect when you have just a single transcript provided as a reliable aid to perception of an indistinct covert recording.


See more examples in our News section, or proceed to the Case Study.

Sign up here for occasional news about Forensic Transcription (no spam!)