TGRWT #10: Pineapple and blue cheese


Back from a trip abroad I can finally relay the announcement of the 10th round of “They go really well together”. The round is hosted by David Barzelay at Eat Foo(d) and the challenge this time is to combine pineapple and blue cheese.

In the list Heston Blumenthal posted on eGullet a long time ago it was pointed out that for this combination a certain level of ketones (molecules with a special carbon oxygen bond) for the combination to work. Since there was no further information I did some further research – and voilà – lists of odorants for pineapple and gorgonzola cheese have been published! But surprisingly there was no overlap between the compounds for which odor activity values (OAV) had been calculated. A possible reason is that only 12 and 15 compounds for pineapple and gorgonzola respectively where quantified so that OAV’s could be calculated (notice that it is the quantification which is really time consuming when doing this kind of research). Another point is that different experimental techniques where used in quantifying the volatiles. But regarding the ketones the gorgonzola article at least shows that “natural” (dry, crumbly) gorgonzola has higher levels of ketones than “creamy” gorgonzola.

To conclude, there is a possibility that the overlap in impact odorants for pineapple and blue cheese has not been uncovered yet, or that this combination can not be explained by overlapping impact odorants (and I should quickly add that this is of course the case for most flavor pairings we encounter in the kitchen!).

Filed under: molecular gastronomy, TGRWT


  1. Trig Says:

    Lesson learnt. Public apology coming this weekend.

  2. Orges Says:

    Dear Martin, it is not that surprising. Probably you remember that we talked about this in Paris this summer. You’re asuming that the flavor pairing theory is totally true, but there is no scientific evidence for it. It is a funny theory for playing with, but there are many examples of foods that do not go together and shares volatiles (and the opposite).


  3. Martin Lersch Says:


    Well – I assume the hypothesis is true, given that there is a sufficient overlap between the impact odorants. One can probably discuss what is “sufficient”, but it’s certainly more than sharing one volatile. As I’ve elaborated on in previous blog posts I belive that OAVs are very helpful in evaluating possible flavor pairings. But unfortunately it’s rather time consuming to obtain those values.

    I totally agree with you (and I’ve stated this several times) that most combinations of food can NOT be explained by this kind of flavor pairing.

    The basis for the TGRWT event is lists of pairings published by Heston Blumenthal and others. AFAIK these are based on the number of volatiles shared, not on odor activity values (OAVs). So necessarily, some (or most?) of the combinations listed (for instance here) can not be explained by the flavor pairing theory. But despite this, some of them still go well together (albeit for reasons unknown).

    The purpose of the TGRWT event is to experimentally test the combinations and come up with suggestions for recipes. Furthermore it is my wish that cooks – if only for a second – consider the fact that food is chemistry and that they are actually eating molecules. I’ve read what the different hosts and participants have written about the TGRWT event, and although most do get the overall idea, only a few are totally into the details I discuss here.

    If we look at the experiments performed, scientifically speaking there are four possible outcomes:

    1) If a combination works AND there is a significant overlap between the impact odorants this strenghtens the flavor pairing hypothesis.

    2) If a combination works, AND there is no data available about the impact odorants OR the data is available, but there is not a significant overlap, this result falls into the same category as most combinations we encounter in the kitchen. The result says nothing about the validity of the flavor pairing hypothesis. What is says is that most flavor pairings can not be explained by the flavor pairing hypothesis.

    3) If a combination doesn’t work AND there is a significant overlap between the impact odorants this might be because the foods were combined in the wrong ratio or that the recipe simply was bad. But it doesn’t tell us more about the flavor pairing hypothesis.

    4) If a combination doesn’t work AND there is no data available about the impact odorants OR the data is available, but there is not a significant overlap, again this could simply be a result of a poor cooking, and so it doesn’t tell us anything about the flavor pairing hypothesis.

    Alternative 1 will strenghten the flavor pairing hypothesis, whereas 2-4 won’t make any difference at all, seen from a scientific viewpoint.

    But from a gastronomic viewpoint both 1 and 2 are interesting, because this means that creative people around the blogosphere have come up with new recipes and dishes that often surprise. And to end – this reminds me of the famous statment made by Brillat-Savarin:

    “The invention of a new dish is of greater importance to the happiness of mankind, than the discovery of a new star”

  4. Kiriel Says:

    Not a new combination to me, pineapple and blue cheese… I guess the challenge will be to find a new dish for the pairing. mmm… fun fun fun!

  5. Martin Lersch Says:


    It’s perfectly OK to participate using a recipe you have found somewhere!

  6. Orges Says:


    It is great to discuss about these topics: flavour volatile compounds have been my area of research for 17 years, and I love to talk about them.

    Nevertheless, I cannot agree with you. When doing scientific experiments (although we know this is not pure science, but a field between science and cooking, or something like this) there is no room for these potential interpretations of the theory you told us. I mean, if the hypothesis does not work (and if the “experiment” is correctly designed) then the hypothesis is not true. Perhaps it would be necessary a deeper research (scientific, of course) to check the hypothesis, but at a first glance, there are many, many other factors that can explain the combinations between foods. Perhaps in some of then the sharing of volatiles could explain something, but in many others don’t. So, it is not so strange that foods sharing volatiles do not going well together, and that food with no overlap goes well together.

    And finally, as I told you, once you have analysed thousands of differnt samples, you realise that many of the most important volatiles for many differnt foods are quite the same, and that the quantity and porportions are really key in explaining the overall flavour. For example, different meats share most impact flavour compounds, and most of the time they do not go well together.