Not baking

Conference consommé

Banana consommé. Pure, clear, unadulterated, liquid essence of banana. (Actually, it’s technically a consommé of roasted, caramelised bananas.) This precious liquid was obtained using a gelatin filtration method, which allows the liquid containing the soluble flavour compounds to be removed while the flesh of the banana remains trapped behind due to the network that the gelatin forms in the mixture.

I bet you never knew how much you needed banana consommé in your life until now! It must have a million purposes! Mix it with… other liquids! Make it into… ice cubes? I don’t know…

But there’s a beautiful story behind this banana consommé’s existence.

When some other lab members and I went to a conference on the frontal lobes of the brain in Toronto in March this year, we were all horrifically jet-lagged. I spent the week experiencing almost constant nausea. We were falling asleep during so many of the talks. Our resident psychiatrist was have lucid dreams in which he saw the current speaker giving harsh scores to children for their figure-skating routines. However, in our semi-conscious state, there was one thing that several of us took away from the conference: banana juice, and the importance thereof. Turns out that macaques (the primates often used in neuroscience experiments) absolutely love banana juice. One of the speakers was saying that macaques love banana juice above all else: apple juice, sultanas, grape juice, whatever, don’t bother, it’s all about the banana juice.

If you want to train a macaque to do a task, use banana juice as a reward. If you want to study the reward circuitry of the brain in macaques, use banana juice (or cocaine… or juice and cocaine, as this study did). If you want to see macaque neurons fire in response to a cue on a computer screen, make sure that cue is usually followed by a drink of banana juice, and then you’ll see particular neurons fire in anticipation of the banana juice.

But what is banana juice? Is it just puréed banana? Because that, my friend, would be banana purée. How does one juice a banana?

These were incredibly important questions in our jet-lagged brains.

Ultimately, it resulted in my decision to make banana consommé, to bring into existence an unadulterated banana essence. This, I decided, was closest one could possibly get to the definition of banana juice. And it was beautiful to behold.

I imagine macaques would gnaw through a brick wall to get to this stuff.

Later on, I decided to follow up my consommé-ing success with another tribute to our Toronto trip: clamato consommé, since clamato juice was served on the plane between Vancouver and Toronto. Clamato juice is clam juice and tomato juice, for the lucky souls who have been fortunate enough to avoid knowing about its existence. None of us had heard of it or tried it before and we all desperately wish we could repress our memories of it now, because our simple minds were not prepared for it.

I couldn’t get any clams for my recipe, so I used mussels. I blended them with tomatoes (almost breaking my 600W stick-blender in the process because of the horrible sinewy tissue of the mussels), did a gelatin filtration, and produced clamato (well, musselato) consommé.

Then I took it to work and made our resident psychiatrist drink it. I think he had tears in his eyes.

I then put the clamato juice in the fridge at work, with a smiley-face on it and I don’t know what happened to it but it did eventually disappear.

Read on for the recipe for banana consommé using gelatin filtration. And feel free to suggest interesting potential uses for banana consommé, because I only really made this as a proof of concept and didn’t think much beyond that…

Darkplace soup

May I tempt you with this alluring and sophisticated dish?

It’s a poached egg in soup.

Used since ancient Greek times to woo one’s objet d’amour, poached egg in soup remains one of the most alluring, enchanting and salacious gastronomical expressions of undying passion. The egg, representing fertility and perched perkily upon the thick molten soup, is like an unblinking bulbous eye within which you glimpse an unmistakeable come-hither look.

Or so I imagine. All I can say is that this is my loving, labour-intense, bizarre, gastronomical tribute to Garth Marenghi’s Darkplace, one of the greatest television shows ever. In the mighty concluding episode of the series, Dr Lucien Sanchez, hotshot surgeon, woos Linda, the woman who is slowly turning into broccoli, using an egg poached in soup. Such a poignant and beautiful gesture… that looks like this:

My variation has nothing on the original, of course, but I have exerted myself and my creativity to produce this dessert version of the egg poached in soup: a strawberry mousse tinged with Grand Marnier and dark chocolate (and a bucket of food colouring — damn, that mousse was a delicate pale pink), a circle of vanilla milk gel, and a yolk made out of a sphere of saffron-flavoured apricot nectar.

And so with this I salute you, Garth Marenghi: author, dream-weaver, visionary, plus actor. You taught me to love, and to convey that love through an egg poached in soup (mayhap with a pork pie or sausage roll). If only there was more of such epic beauty in the world.

(Darkplace Episode 1 Part 1 here.)

As for the neuroscience for this post, just a quick mention of something very interesting I came across when reading up on lexical-gustatory synaesthesia, when people experience particular tastes in response to hearing, reading or thinking particular words (see my previous post if you haven’t already).

We’re all familiar with that indecently enraging thing known as the tip-of-the-tongue phenomenon. You’re trying to think of a word and YOU. JUST. CAN’T. BUT. YOU. HAVE. ALMOST. GOT. IT. IT’S ON THE TIP OF MY TONGUE. OH. GOD. I’M SO ANGRY. This is a pretty difficult thing to research — everyone will have different words that cause this phenomenon for them (compromise and renege do this to me endlessly for some reason, and yes, they did it while I was writing this, gaaarrghhh), and once you’ve reminded someone of the word then there’s no point in testing them with that word again any time soon.

Apparently researchers have worked out some words that are pretty good at generating the phenomenon. For example, “What’s the name of the navigation instrument you use to measure the angle between two objects?” — the answer is a sextant, which is apparently a difficult word to recall (I had no problem recalling when my psychology lecturer asked that in one of my undergraduate courses, and it’s not like I’ve ever so much as held one, although at least half of the class did have trouble recalling the name). Another one that often generates tip-of-the-tongue is platypus, apparently. That’s really not going to work for anyone who grew up in Australia, that’s for sure — I think most 5-year-olds here could name a platypus without hesitation.

Anyway, the tip-of-the-tongue phenomenon occurs when you can access the word’s meaning from your memory, but you can’t quite remember its actual sound.

The particularly interesting part is when you try to generate a tip-of-the-tongue state in a lexical-gustatory synaesthete. Why?

Because they experience the taste before they can recall the word.

So for instance, one participant, MK, was shown a picture of castanets and was asked to say what they were. She couldn’t quite remember the name, but she did experience the taste of tuna while trying to figure out the name of what she was looking at. Then, later on, when she was asked what taste is usually associated with castanets, she confirmed it was tuna. And just to be extra thorough and make sure she wasn’t just saying that castanets was associated with the taste of tuna because that’s what the researchers would like to hear, a year later, the participants in this study were given a surprise re-test and asked which tastes were associated with which words. And yep, MK again said that castanets was associated with tuna.

This tells us something pretty interesting about lexical-gustatory synaesthesia, then: it’s not the actual phonological sound of the word that triggers the associated taste, but its meaning, since the tip-of-the-tongue phenomenon occurs when you can dredge up the word’s meaning from your memory but can’t quite get the word itself.

Which makes it all the more odd that these synaesthetes experience tastes in association with made-up, non-existent words, like keach or spluk (like I mentioned in this post). Maybe there are different levels of processing going on in lexical-gustatory synaesthesia? One involving the meaning of words and one involving their orthographic and phonological aspects, and these levels of processing can sometimes work independently of each other? Maybe the initial association between the word and a taste is based on orthographic or phonological aspects of the word, but then the association becomes so strong and automatic that the taste is conferred to the word’s meaning instead? I don’t know, but there is the potential for some very interesting and elegant studies to be formulated to answer those questions!

And now, what you really wanted: the recipe for fake egg poached in soup, brought to via a brief detour through the outer regions of molecular gastronomy land.

Read on for the recipe for fake egg poached in soup — strawberry mousse, vanilla milk gel and saffron apricot nectar sphere.

Hot ice-cream success! (And synaesthesia.)

You might remember my previous attempt at making hot ice-cream — the ice cream that you poach in hot water and then it melts as it cools to room temperature. It worked, technically, but I wasn’t happy with the flavour since it just tasted like cream cheese, which, you know, doesn’t taste like ice-cream (unless you have cream cheese ice-cream, I guess, but who has that? That’s right, no one. Well, some people, but they’re not the sort of people I want anything to do with).

So I went on the quest to develop a hot ice-cream recipe that tastes like normal vanilla icecream. And after some experimentation and various mishaps, ye gods, I think I’ve actually done it.

Jack at Homologous Legs happened to tweet about the old recipe at one stage, saying that he’d be all for it if the hot ice-cream was in iced chocolate, as opposed to the iced coffee I’d put it in to make the reverse affogato. And I thought yes, that’s a good point. So in this iteration of the recipe, I changed the lovely bath in which the hot ice-cream resides from iced coffee to a cinnamon chocolate frappé. Just because I can. (But methylcellulose also thickens at low temperatures as well as 50°C+ temperatures, so the frappé is going to cause the ice-cream to melt and then solidify again, which I pretend is all part of the thrilling and dynamic dining experience.)

So there you go. Hot ice-cream that melts as it cools. It’s kind of amazing and it defies the senses.

You know else is kind of amazing and defies the senses? (Behold the world’s most stunning segue.)


For those of you who have the unfortunate circumstances of not knowing about the mind-melting amazingness of synaesthesia, it is a condition in which perception of a stimulus in one modality is experienced in an additional modality. What does that mean?

Well, imagine that every time you heard a dog bark, your visual perception was spontaneously and involuntarily tinged with purple.

Or imagine that as you read this blog, you’re visually experiencing a relentless cavalcade of colours with each different letter you look at.

That happens to synaesthetes — they experience a vivid and completely involuntary cross-over between senses (and possibly cognitive systems). And there are different kinds of synaesthesia involving cross-over between different senses and cognitive processes such as language processing. Grapheme-colour synaesthesia is the one where people experience a colour in association with different letters or numbers (e.g. the letter A conjures up the perception of red or is perceived as being inherently red). Sound-colour synaesthesia is the one where hearing a particular noise or even a particular note on a musical instrument causes the synaesthete to experience a particular colour. And studies have shown that these particular associations (such as A = red or dog barking = purple) generally remain steady over the synaesthete’s lifetime. So play a B-flat on a cello for a synaesthete and perhaps they see orange, then come back 25 years later and play that B-flat on a cello again and ask them what colour they experience — it’ll be orange.

But within the scope of the neuroscience of taste perception, one type of synaesthesia is just lobe-explodingly interesting: lexical-gustatory synaesthesia. When people with this synaesthesia hear or read or even think a particular word, they consequently experience a particular taste. Not just that, but words conjure up the sensation of taste and temperature and texture in the mouth (perhaps not surprising to those of you who remember when I discussed the primary taste cortex in this post and mentioned that some neurons in the primary taste cortex fire preferentially to particular food temperatures and textures as well as tastes).

It’s estimated that less than 0.2% of the population experiences this lexical-gustatory synaesthesia, so it’s pretty uncommon and it’s difficult to research when there are so few potential research participants (damn it, I find it hard to get research participants and all I need are right-handed males aged 18-40). But here are some of the word-taste associations that lexical-gustatory synaesthetes experience, according to the literature…

  • “for synaesthete JIW, the word jail floods the mouth with the taste of bacon”
  • “for SKM, the word dean evokes the precise and consistent notion of minced beef in gravy”
  • “for synaesthete JG for example, the name Adrian tastes of lettuce coated with Caesar salad dressing, and for CS, part tastes of chicken noodle soup”
  • “for JIW, the words message, college and edge all taste of pork-pie/sausage meat”
  • tambourine is very crumbly biscuit”
  • “for synaesthete JIW, words containing /k/ tend to taste of egg/yolk, e.g. accept, York, chuck, fax
  • “for example, Alessandro tasted of ‘fried potatoes’ and gave the smell of ‘burnt wood’”.
  • “for example, for JIW, the non-word spluk tastes of yoghurt, and so does its near-neighbour luck
  • “for LAS, for example, beef tastes of ‘horrible overcooked, dried-out beef’”
  • “for JIW, the word this tastes of bread soaked in tomato soup, while the name Philip tastes of unripe oranges”
  • “for participant GS, the word clue tastes of ‘cold, leftover potato scraps with congealing pot-roast gravy’”

Pretty specific, huh? And kind of, well, unpleasant in some cases. I’m in no hurry to taste congealing pot-roast gravy, never mind have it forced upon me by my brain every time I read Nancy Drew (I’ve never read Nancy Drew, but what if I wanted to? What if I wanted to?!).

You might have guessed from those little excerpts the direction in which lexical-gustatory synaesthesia research has headed — what are the specifics of a word that mean it’s associated with a taste? Are there generalisable trends within each synaesthete, e.g. do words that sound the same generate similar tastes? And what about made-up words — do they generate a taste? And if they do, why? More, tell me more!

Well, to start with, one thing is pretty straight forward — if something has a taste, the word for that something will generally taste like that something. So the word asparagus conjures up the taste of asparagus. It might be the taste of over-cooked asparagus for one synaesthete and raw asparagus with freshly ground pepper for another, but generally there is a pretty blatant link between food words and their respective food tastes.

Another of the most clear-cut findings is that sometimes there are obvious associations between parts of words and the name of what they taste like. For example, “cinema tastes of ‘cinnamon rolls’ (JG), Jackson tastes of ‘Cracker Jacks’ (both MZ and CS), village tastes of ‘vanilla slice’ (SKM), and dogma tastes of ‘hotdogs’ (DMS).” So there’s some sort of word-association going on here based on key sounds and letter groups within a word.

It also seems like particular sounds within words might be the point that the taste association clings to. That’s why message, college and edge taste of sausage meat (the /dzj/ sound in all of them and in sausage) and things with a /k/ sound taste of egg yolk to JIW.

This all kind of makes sense, but what would you predict would happen with made-up words, non-words? These do generate tastes in the synaesthete, even if the synaesthete has never encountered them before. It looks like non-words associate with a taste based on how similar they are to real words. For example, if presented with the non-word keach, a lexical-gustatory synaesthete might experience the taste that’s usually associated with beach or teach. Even non-synaesthetes can see the similarity between those words, but synaesthetes might have a broadened sense of word similarity, in that a word like spluk results in the taste that’s usually associated with luck, even though spluk and luck are technically very different words.

Despite there being very few individuals with lexical-gustatory synaesthesia to study, there are a few papers out there on the topic that suggest some very interesting theoretical underpinnings for this type of synaesthesia, and some incredibly nifty experiments have gone down. I will, however, go into that another time (well, in my next post).

In the mean time, just reflect on the fact that you are not currently tasting lumpy gravy or cold potato skins or oily salad dressing while reading this. If you were anything like the synaesthete JIW, you would be experiencing tastes in response to 56% of words from your own speech, the speech of others, inner speech and reading. An almost constant parade of tastes…

Personally, my strategy would be that I’d memorise a list of all the words that don’t generate a taste, then try to speak and think exclusively in those words, just for a little respite. Think of it as being like that guy who wrote a story that didn’t contain the letter E. So… difficult… I think… hi… how do you do today… do you… want to… stuff into your mouth… hot rich dairy product? Look down for… instructions…

Read on for the recipe for hot vanilla ice-cream in cinnamon chocolate frappé.

Beetrooccino & a certain xanthine alkaloid

Oh, caffeine. Or as its ultra-catchy IUPAC name would have it, 1,3,7-trimethyl-1H-purine-2,6(3H,7H)-dione. It’s the most widely consumed psychoactive substance in the world, and I’m consuming it right now, while I type! Just think of the things that have possibly been written under the influence of caffeine. Probably, like, War & Peace and stuff. And so this blog post pompously insinuates itself into that mighty company.

But how much of our dependence upon (and love affair with) caffeine is the result of its chemical effects on our brains, and how much is just in the mind? (Figuratively.)

It’s more than a little bit in the mind. You should know that by now.

One study had a little look at what happened to people’s ability to pay attention and detect important information after they consumed either a caffeinated or non-caffeinated drink, depending on whether they were told about its correct identity or misled (that’s right, another experiment full of blatant lies — I love it).

So all participants went through four sessions:

  • One in which they were given a caffeinated drink and told it was caffeinated (truth!).
  • One in which they were given a caffeinated drink and told it was non-caffeinated (lie!).
  • One in which they were given a non-caffeinated drink and told it was non-caffeinated (truth!).
  • One in which they were given a non-caffeinated drink and told it was caffeinated (lie!).

Different participants did the different sessions in different orders. They didn’t know that sometimes they were being lied to; they’re just drinking these drinks and then doing a computer task. So in each session, a little while after they consumed the drink, they performed a computer task in which numbers flashed up on the screen and participants had to pay attention and try to spot a target pattern (e.g. when the same number flashes up twice in a row and is an even number). This was a measure of their vigilance.

What would you expect would happen? We consume caffeine usually because we want to be more alert and pay attention, so having caffeine should improve vigilance, right? Well, yes, caffeine did improve vigilance.

But only when participants were told that their drink had caffeine in it.

If they had the caffeinated drink but were told it was decaffeinated, they performed pretty much exactly the same as when there really had been no caffeine in their drink. So their expectation that there was no caffeine to improve their performance meant that their performance wasn’t improved, even when the caffeine was there to act chemically on their brain.

However, this expectation effect wasn’t there for the non-caffeinated conditions: people performed pretty much equally after a non-caffeinated drink regardless of whether they thought it was caffeinated or not.

So what can we take away from this? Subtle relationships, people. Subtle relationships. It seems that there is an interaction between the chemical effects of caffeine and our normal expectations of what caffeine is going to do to us. The caffeine needs to be there, but we also need to expect it to be there and to work for it to actually work.

Or, more concisely, you can take this message away from it:


And with that in mind, fancy some beetroot coffee? Or as I call it… beetrooccino?

Yeah so this is pretty much inspired by Heston Blumenthal’s lobsterccino, but (1) I’m not aiming to emulate the crazy excess of the 1980s like Heston was (I wasn’t there for half that decade), (2) beetroot and coffee share odour compounds that might make them complement each other rather well and (3) I was buying molecular gastronomy supplies online and my order was under the minimum order value by 5 cents, so I bought a $4 packet of beetroot powder and had to figure out something to do with it.

And the beetrooccino is good. It is actually really good. The coffee and beetroot do complement each other very well, although I have to say it is rather a savoury drink. I wouldn’t want to finish a meal with this, but… it might work as a slightly crazy hors d’œuvre prior to food.

Read on for the recipe for beetrooccino…

Srs nurrosiense tiem

Have you ever been walking down the street when you stopped dead in your tracks and thought “OH GOD OH GOD HOW DO I TASTE THINGS?!! HOW DOES MY BRAIN LET ME TASTE SALTY PRETZELS UNNNGHHHH OH GOD I DON’T KNOW FFFFFFFFFFFFFWHAT DO I DO?”. I bet you have. I haven’t, but that’s because I do know (in a very rough approximation), although I imagine that if I didn’t know, I would be shrieking and my eyes would be bulging out of their sockets and I would possibly be convulsing in the middle of a road somewhere. Such is life.

So that’s it. It’s neuroscience time. None of this namby-pamby psychology stuff with people reading lists of words and acting ever so slightly differently as a result. We’re getting on a little nano-rocket and riding into the neuron metropolis. I’m a neuroscientist and I’m bringing my A-game. Are you? Yes? Is it folded up in your backpack? Ok.

Right then. After your tastebuds have detected food and the basic information about the food has been transmitted along nerves to the more fancy parts of the brain, what does the brain do? How do the actual neurons in the brain respond, and what sort of information do they respond to?

I’m going to concentrate on just one part of the brain for now – the primary taste cortex, consisting of sections called the insula and the frontal operculum. This primary taste cortex is place where the brain starts to integrate all the different bits of information about food. After this complex processing has begun, the primary taste cortex interacts with a huge range of other brain areas that are involved in the processing and perception of vision, smell and touch (not surprising, given the multisensory experience that is eating).

The other senses also have primary cortices where this sort of processing of sensory information goes on at a higher level, and taste perception shows some striking similarities to perception in the other senses.

For instance, in the primary visual cortex, there are neurons that only fire when you see very particular things. For example, there are neurons that fire in response to a vertical line in your field of vision (the side of house, a flagpole, etc) but don’t fire or decrease their firing below normal in response to a horizontal line (the horizon, the top of a desk, etc), and likewise, there are some neurons that only fire in response to horizontal lines and they’re not so responsive when presented with vertical lines. Neurons can be very specifically tuned and only respond to a very distinct stimulus, or they can be more general and fire in response to a range of stimuli. The same is the case in the primary taste cortex.

In the primary taste cortex, neurons respond to lots of different properties of food. Verhagen and colleagues looked at neurons in the primary taste cortex and how these neurons fired in response to food of varying taste, temperature, grittiness, viscosity and fat texture. They found that neurons can have a very specific profile of stimuli that they actually fire in response to.

53% of the neurons tested fired in response to the thickness or viscosity of food. When the researchers tested different viscosities (using carboxymethylcellulose), some neurons fired more in response to very thick consistencies whereas others fired in response to a runnier consistency.

8% of the neurons tested responded to grittiness as a consistency. So I guess these neurons will be firing like mad if you eat porridge with sand in it.

Another 8% of neurons tested actually responded to fat, but they identified fat by its texture, not by any chemical method. The researchers figured this out because the same neurons fired in response to non-fat oils that had the same texture as the fats tested.

35% of neurons tested responded to the temperature of whatever was in the mouth. The researchers used water at different temperatures (10°C for a chilled drink, 42°C for a warm drink, 37°C for body temperature and 23°C for room temperature) and some neurons fired more in response to particular temperatures than others.

6% of the neurons tested responded to capsaicin, the hot compound of chilli peppers. The interesting thing here is that these neurons didn’t respond to the warmest water temperature (42°C). So even though capsaicin is experienced as heat, it didn’t result in firing of the neurons that had fired to a warm liquid. However, it turns out that 42°C might have just been an unlucky temperature to choose as the upper limit in the experiment, as capsaicin’s hot effect is achieved through a particular type of receptor that only responds to temperatures greater than… 43°C. So if a hotter liquid had also been used in the experiment, then maybe neurons would have been found that fired in response to the hot liquid and capsaicin.

And of course, a fair few neurons responded to taste. 56% of neurons tested responded to taste, which was tested using blackcurrant juice (for sweetness), table salt (for saltiness), weak hydrochloric acid (for sourness), quinine (for bitterness) and MSG (for umami).

Now, these neurons didn’t always just fire preferentially for taste or temperature or viscosity or whatever — about half of them fired in response to combinations of these classes of stimuli. For example, 23% of them fired in response to both taste and temperature. A couple of them fired in response to taste and temperature and viscosity and fat. So there was overlap between the groups of neurons involved in responding to the different properties of the food, which allows for a much more complex and nuanced representation of food in the brain.

The interesting thing was that these neurons in the primary taste cortex did not fire in response to odor or the sight of food. This tells us that it isn’t until a later stage that taste information and visual and olfactory information get integrated. And that is indeed what the pathways of sensation in the brain tell us (as seen in the diagram below that I hastily knocked together). It’s rather complicated but hey, that’s the brain for you:

So don’t worry about understanding this whole mess (simply stand back and appreciate the complexity of that soft lumpy thing inside your skull). Just know that arrows indicate the flow of information, and blunt-ended lines indicate an inhibitory effect where the flow of information is dampened. And you can see that in the pink taste pathway, by the time information has gotten to the primary taste cortex from the taste receptors in the tongue, it hasn’t interacted with any of the other pathways. But in the next step, when information flows into the orbitofrontal cortex and amygdala, it has its first chance to interact with information from the other sensory modalities, vision (green), smell (gold) and touch (blue).

The orbitofrontal cortex is also the part of the brain responsible for the representation of the palatability and pleasantness of food, which means that how enjoyable food is probably results from an interplay of visual, taste, olfactory and touch factors. So no matter how delicious that pie tastes, if it happens to look like horse manure then the orbitofrontal cortex just isn’t going to assign as much of a pleasant experience to it.

So there you go. The basics of how the brain processes taste. But as is always the case, it’s a whole lot more complicated than that. When Verhagen and colleagues were looking at the firing properties of neurons in the primary taste cortex, they only reported on the neurons that responded to at least one of the stimuli in the experiment, whether it be a particular taste or viscosity, the presence of grittiness or fat, water of a particular temperature, and so on. In all, they tested 29 different stimuli to see if neurons fired in response. They found 62 neurons that did this. They found 1,060 that didn’t.

What are these other 1,060 neurons doing? What are the other hundreds of thousands of neurons that weren’t tested in the primary taste cortex doing? What do they respond to? What’s their job?

The science continues!

And in the meantime, while you wait, have a HobNob shake. It’s based on the recipe for the Max Brenner cookie shake, lord among shakes that it is. It looks good, it tastes good, it smells good and… well I guess it has a good texture, in the scheme of things. Your orbitofrontal cortex is going to assign awesomeness to this via the striatum so hard.

Read on for the recipe for a HobNob shake.