Nevaeh ot Yawriats

I just remembered hearing this a while ago at a public talk given by Simon Singh. I guess many of you will have come across it before, but there’s no harm in repeating it. I don’t know why it popped into my head at this particular moment, but perhaps it’s because I’ve been reading some stuff about how my colleagues in gravitational wave research use templates to try to detect specific patterns in noisy data. The method involves cross-correlating a simulated signal against the data until a match is obtained; the problem is often how to assess the probability of  a “chance” coincidence correctly and thus avoid spurious detections. The following might perhaps be a useful warning that unless you do this carefully, you only get out what you put in!

This is an excerpt from the classic  track Stairway to Heaven, by the popular beat combo Led Zeppelin, played backwards. I suggest that you listen to it once without looking at the words on the video, and then again with the words in front of you. If you haven’t heard/seen  it before, I think you’ll find it surprising…

Of course the proper way to interpret (or dismiss) matches like this is to use tools based on  Bayesian inference….

10 Responses to “Nevaeh ot Yawriats”

  1. Any other kind of inference would be bayseless.

  2. telescoper Says:

    • Anton Garrett Says:

      Playing this constitutes an instant aural explanation of what is really going on with Stairway to Heaven. Subsequent rock bands have gone in for deliberate backward masking, but it is very obvious that Led Zep didn’t.

      It is very hard to apply Bayesian analysis when the list of possibilities is not known. In that situation I would apply neural net pattern recognition and not try to get probabilities.

      • telescoper Says:

        Quite, but when you have a well-defined bank of templates and a good understanding of the noise characteristics then I think a full Bayesian approach is manageable (although by no means easy)…

  3. “Subsequent rock bands have gone in for deliberate backward masking, but it is very obvious that Led Zep didn’t.”

    Agreed, but most (all?) backward masking was a jokey reference to the claims about subliminal influence etc. As one defendant at a trial said, if it worked we would put in “buy more of our records” not tell our fans to kill themselves.

    I haven’t seen the clip yet, but somewhere on the intertubes is an hour-long (or so) video of a presentation on this topic, using the Stairway to Heaven passage as an example.

    • Anton Garrett Says:

      I couldn’t agree more that such concealed backward messages are not picked up subliminally.

      At a guess, the clip on the subject is by a former rock musician, now churchman in California, called Pastor Joe Schimmel. So far as I am concerned his heart is in the right place but he makes some mistakes in his analyses of specific songs.

  4. Yes, I saw this in a ‘skeptic’ talk too, though not one of Simon Singh’s – might have been Steve Novella.

    It is really striking how you can ‘hear’ the word once you’ve seen written down what you’re ‘supposed’ to hear – it works even if you’re not told ‘they’re definitely there’. As a demonstration of the brain’s power to fool itself via its being programmed to find patterns it would be hard to better.

  5. Anton Garrett Says:

    Relevant and hilarious:

  6. Dave Carter Says:

    A tool shed, are you serious? I must say I heard not one word of that without the words there in front of me.

  7. Chris Morris exposed the evils of backmasking a couple of decades ago in an important and influential On The Hour report.

    Oh, Rolf, where did it all go wrong….

Leave a Reply

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

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

Google photo

You are commenting using your Google 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 )

Connecting to %s

%d bloggers like this: