Philip Ball is surprised to discover just how sensitive we are about our feelings

Chemistry is the most sensuous science. Vision, taste and smell have always been among the chemist’s key analytical tools. It was common even into the last century for some researchers to dab new compounds onto the tongue to gauge their properties, and of course chemistry’s smells are both useful and notorious. But how exactly do we make these sensory judgements? 

The perceptual dimensions of colour are relatively simple and well-explored – typically expressed as hue, brightness or tone, and intensity or saturation. Smell is trickier. The fragrance industry has naturally examined the issue in depth, but the conclusions remain anecdotal. The ‘odour maps’ of perfumiers are deliciously evocative, with descriptors such as ‘woody’, ‘citrus’, ‘floral’ and ‘aldehydic’ radiating from a central focus – but it’s not clear what the basic cardinal directions are in such a classification.1 Taste might be mapped via the five fundamental dimensions of sweet, sour, bitter, salty and umami, but not only does that fail to capture the aromatic dimensions of flavour, it also neglects the importance of texture. Besides, the relationship of taste to chemistry is unreliable: sweetness doesn’t necessarily correlate with nutritional value (think of saccharin), nor bitterness with toxicity.2

Of all the senses, touch is probably the least understood

In such ways, the instruments through which we experience the world are not like the instruments of science, which tend to measure a particular, well-defined quantity such as wavelength or chemical shift. We sense qualities more than quantities, perhaps without quite being able to say what those qualities are. In music, timbre is one of the most emotionally salient aspects of sound and yet we don’t know what the basic dimensions of timbre are, or if indeed it has any.3

Touching thought 

Of all the senses, touch is probably the least understood in its subjective features. We know something about the perceptual limits – it becomes almost impossible, for example, to sense by touch any static object smaller than 0.2mm across.4 But texture is more qualitative. Some studies have indicated that we judge it along hard–soft and smooth–rough axes, and perhaps sticky–slippery.5 But the question is whether these sensations can be matched to any objective, physical property of a surface.

One of the challenges in studying that issue is to prepare samples that can be reliably compared – that is, which differ only in some single, quantifiable property. A new study by Mark Rutland of the KTH Royal Institute of Technology in Stockholm, Sweden, and his coworkers addresses that problem.6 The group has prepared stiff polymer films with a series of parallel sinusoidal ridges that differ only in their wavelength (300nm to 90mm) and amplitude (7nm to 4.5mm), by inducing wrinkles in a carefully controlled fashion. The films, of a light-curable resin, are moulded on a template of flexible polydimethylsiloxane (PDMS), which is stretched and oxidised so that the surface layer becomes stiffer, creating a bilayer. Releasing the strain then produces wrinkles whose amplitude and wavelength are determined by the initial stretching and degree of oxidation.

Can you feel it?

Rutland and colleagues asked human subjects to feel these surfaces with their index finger, brushing perpendicular to the ridges, and judge the differences between pairs. They found that these evaluations could be plotted on a two-dimensional space, and that the two dimensions seem to correlate with objectively quantifiable parameters, namely the finger friction coefficient and the ridge wavelength. The relationships are not linear, however, but sigmoidal: in other words, the subjects lose discrimination at particularly large or small values of the two variables.

What’s more, these variables seem to be analogous to the rough–smooth (wavelength) and sticky–slippery (friction) distinctions suggested in earlier studies. There’s no variation of hardness in these experiments, since all the surfaces are made from the same stuff.

All this helps to pin down what determines tactile texture. But there’s a further striking aspect of the work. Some subjects were able to register a tactile difference between surfaces patterned with grooves just 13nm high and ‘smooth’ surfaces. It seems remarkable that fingertips, which are themselves of course grooved with macroscopic ridges, can sense phenomena at the nanoscale. Yet that finding isn’t unprecedented. Our fine touch response is mediated by mechanoreceptors called Pacinian corpuscles: nerve endings activated by pressure-sensitive sodium channels. These can register a neural response to vibration amplitudes of just 10nm applied to the skin.7 All the same, probing this sort of surface topography is more generally associated with the atomic force microscope (AFM), and indeed modifications to that device have previously been introduced to give it a haptic interface. Yet it seems that, at least to some degree, the finger can already compete with the AFM’s delicate needle tip.