The visual neuroscience literature is rife with two conflicting claims, made simultaneously: 1. Attempts to produce repeatable correlations between individual neurons and experimental conditions consistently fail; activity is variable in the extreme, even to identical stimuli. 2. Authors routinely claim to have ascertained the “tuning preferences” of neurons (on the basis of which they proceed to their experiment). The second claim is a necessary correlate of the logically untenable but persistent notion that neurons act as “feature detectors,” responding most vigorously to their “preferred” stimulus feature. Failure to demonstrate such feature-detector behavior is obviously fatal to the position; but there is no other available in the depauperate intellectual landscape of mainstream vision science. So investigators pretend to ascertain tuning, even as they acknowledge its impossibility. And they don’t pretend to do it only in the case of identical stimuli, but for whole classes of stimuli (potentially infinite) representing some abstract feature, like orientation, or even some non-feature, like “spatial frequency,” i.e. the product of a Fourier analysis the visual system is supposed to perform, for which there is no credible empirical evidence, no conceivable purpose or mechanism.
It’s done in a couple of ways.
Some authors simply assert that they’ve done it – for example: For each isolated neuron, we determined the optimal direction, spatial and temporal frequency, position and size…(Smith, Majaj & Movshon, 2005). We just did it, trust us, the method isn’t important. This is the first of the two categories of articles I’m listing below.
Another approach is to describe your method, vaguely, and to pretend that the results of a particular set of measurements carries special weight (i.e. to ignore the huge variance problem). Britten, Shadlen, Newsome & Movshon (1992), for example, tell us they listened to an audio monitor or eyeballed rasters and made “qualitative” judgments of neuron preferences for speed and direction. Others may describe their method more fully, like Bondy et al (2018). Averages are taken over whatever time intervals the authors psychic advisor suggests (e.g. “When dots were used as the visual stimulus, they were presented for 2 sec, and the average firing rate over the entire period was used as the response metric” (Groh, Born & Newsome (1997)). Articles in this category are in the second list.
The second approach is no better than the secretive “we just did it” ploy. Given that the measurements are known to be highly unreliable, any given sample of “highest response” (even if highest response were a valid notion of neural “coding”) to a given situation at a given moment carries no weight. Note too that there is no claim of a threshold firing rate; “highest response” is only a relative value.
(Readers may get a taste of how variable responses to identical stimuli are from Figure 1 in Taouali et al (2015).)
When experimental data are interpreted, post hoc, based on weak correlations, authors refer to the resulting algorithms as “predicting” the data, even though, of course, the “predictions” only apply in retrospect and have no actual predictive power. The known unreliability of the kind of measurements serving as premises for these experiments (“we presented stimuli in neuron’s preferred direction” etc) makes use of the term even more inexcusable.
The stark contrast between the two empirical claims (that measurements are reliable in theory and practice, vs they are unreliable in practice) is perhaps the clearest illustration of the cargo cult nature of contemporary neuroscience. Bear in mind also that most if not all (probably all) of the articles employ significance levels of p< .05.
I. We just did it school
Newsome, Britten & Movshon (1989) Neuronal correlates of a perceptual decision Nature.
Salzman, Britten & Newsome (1990). Cortical Microstimulation Influences Perceptual Judgments of Motion Direction. Nature.
Newsome, Britten, Salzman, & Movshon (1990) Neuronal Mechanisms of Motion Perception/ Cold Spring Harbor Symposia on Quantitative Biology
Nicole C. Rust, Odelia Schwartz , J. Anthony Movshon , Eero P. Simoncelli (2005) Spatiotemporal elements of macaque v1 receptive fields Neuron.
Smith, Majaj & Movshon (2005) Dynamics of motion signaling by neurons in macaque area MT Nature Neuroscience.
Nicole C Rust , Valerio Mante , Eero P Simoncelli , J Anthony Movshon (2006) How MT cells analyze the motion of visual patterns Nature Neuroscience.
Smith, Zhan & Cook (2011) The Functional Link between Area MT Neural Fluctuations and Detection of a Brief Motion Stimulus. Journal of Neuroscience.
Jazayeri, Wallisch, & Movshon (2012) Dynamics of macaque MT cell responses to grating triplets Journal of Neuroscience.
Robbe L T Goris , J Anthony Movshon , Eero P Simoncelli (2014) Partitioning neuronal variability Nature Neuroscience.
Vintch, Movshon, Simoncelli (2015) A Convolutional Subunit Model for Neuronal Responses in Macaque V1 Journal of Neuroscience.
Hashemi, Golzar, Smith & Cook (2018) The Magnitude, But Not the Sign, of MT Single-Trial Spike-Time Correlations Predicts Motion Detection Performance Journal of Neuroscience.
II. We did it our way school
A. Mikami , W. T. Newsome , R. H. Wurtz *1986) Motion selectivity in macaque visual cortex. I. Mechanisms of direction and speed selectivity in extrastriate area MTJournal of Neurophysiology (1986)
Britten, Shadlen, Newsome & Movshon (1992) The analysis of visual motion: a comparison of neuronal and psychophysical performance Journal of Neuroscience
(The experiments reported in Britten et al (1992) were mined or re-described to produce at least six additional articles, all of which consequently suffer from the same flaws. These are:
William T. Newsome , Kenneth H. Britten , J. Anthony Movshon Neuronal correlates of a perceptual decision Nature (1989) (This is in the first list, since it was only later that the method was reported, indirectly, via Britten et al (1992)).
W.T. Newsome , K.H. Britten , C.D. Salzman , J.A. Movshon Neuronal Mechanisms of Motion Perception Cold Spring Harbor Symposia on Quantitative Biology (1990) (Also in the first list; see comment above).
K H Britten , M N Shadlen , W T Newsome, J A Movshon Responses of neurons in macaque MT to stochastic motion signals Visual Neuroscience (1993)
K. H. Britten , W. T. Newsome , M. N. Shadlen , S. Celebrini , J. A. Movshon (1996) A relationship between behavioral choice and the visual responses of neurons in macaque MT Visual Neuroscience
MN Shadlen , KH Britten , WT Newsome , JA Movshon (1996) A computational analysis of the relationship between neuronal and behavioral responses to visual motion Journal of Neuroscience.
Zohary, E. Shadlen MN & Newsome WT (1994) Correlated neuronal discharge rate and its implications for psychophysical performance Nature.)
Groh, Born & Newsome (1997) How Is a Sensory Map Read Out? Effects of Microstimulation in Visual Area MT on Saccades and Smooth Pursuit Eye Movements Journal of Neuroscience.
James R. Cavanaugh , Wyeth Bair , J. Anthony Movshon (2002) Nature and interaction of signals from the receptive field center and surround in macaque V1 neurons Journal of Neurophysiology.
Graf, Kohn, Jazayeri, Movshon (2011) Partitioning neuronal variability. Nature Neuroscience.
Sabbah, Gemmer, Bhatia-Lin, Manoff, Castro, Siegel, Jeffery & Berson (2017) A retinal code for motion along the gravitational and body axes. Nature
The functional organization of cortical feedback inputs to primary visual cortex Nature Neuroscience
Adrian G. Bondy, Ralf M. Haefner , Bruce G. Cumming (2018) Feedback determines the structure of correlated variability in primary visual cortex Nature Neuroscience
Qing-fang Zhang , Hao Li , Ming Chen , Aike Guo , Yunqing Wen , Mu-ming Poo (2018) Functional organization of intrinsic and feedback presynaptic inputs in the primary visual cortex Proceedings of the National Academy of Sciences
Shushruth, Mazurek, Shadlen (2018) Comparison of Decision-Related Signals in Sensory and Motor Preparatory Responses of Neurons in Area LIP Journal of Neuroscience
Mateo Vélez-Fort , Edward F. Bracey , Sepiedeh Keshavarzi , Charly V. Rousseau , Lee Cossell , Stephen C. Lenzi , Molly Strom , Troy W. Margrie (2018) A Circuit for Integration of Head- and Visual-Motion Signals in Layer 6 of Mouse Primary Visual Cortex, Neuron
Chen, Bao and Tjan (2018) Contextual-dependent attention effect on crowded orientation signals in human visual cortex. Journal of Neuroscience.
Xiaoyang Yao , Jon Cafaro , Amanda J. McLaughlin , Friso R. Postma , David L. Paul , Gautam Awatramani , Greg D. Field (2018) Gap Junctions Contribute to Differential Light Adaptation across Direction-Selective Retinal Ganglion Cells Neuron
Johnston, Seibel, Darnet, Renninger, Orger & Lagnado (2018) A retinal circuit generating a predictive code for orientated features. Biorxiv.
Saleem, Diamanti, Fournier, Harris & Carandini (2018) Coherent encoding of spatial position in visual cortex and hippocampus. Nature.
Kutter, Bostroem, Elger, Mormann & Nieder (2018) Single neurons in the human brain encode numbers. Neuron.
DiPoppa, Ranson, Krumin, Pachitariu, Carandini & Harris (2018) Visual and locomotion shape the interaction between neuron types in mouse visual cortex. Neuron.