A Sensorimotor Numerosity System

Behavioral studies show that humans and other animals can estimate the numerosity of both external stimuli self-produced actions.Recent psychophysical using motor-adaptation techniques have characterized link between action magnitude perception, advancing concept a sensorimotor system encoding internally generated actions.The might reside in parietal cortex. Incoming sensory input provides information for planning execution actions, which yield motor outcomes are themselves inputs. One dimension where perception strongly interact is perception. Many non-human approximately number elements as well their own neurons been identified respond to both. Recent adaptation on also provide evidence neural mechanisms responding externally events actions. Here we advance idea these strong connections may arise from dedicated brain, part more generalized interfacing with processing quantitative magnitudes such space time. Early physiological suggested cerebral cortex organized many areas, each functionally specialized analyze different features perceptual environment [1.Zeki S. A Vision Brain. Blackwell Scientific, 1993Google Scholar,2.Zeki S.M. Functional specialisation visual rhesus monkey.Nature. 1978; 274: 423-428Crossref PubMed Scopus (577) Google Scholar]. work, however, beginning very interactions features, pointing shared associative mechanisms. Perhaps most prominent proposal Theory Of Magnitude (ATOM) model [3.Walsh V. theory magnitude: common cortical metrics time, quantity.Trends Cogn. Sci. 2003; 7: 483-488Abstract Full Text PDF (1207) Scholar], positing human processes jointly about space, optimize Scholar,4.Bueti D. Walsh The representation magnitudes.Philos. Trans. R. Soc. Lond. Ser. B Biol. 2009; 364: 1831-1840Crossref (444) This has spawned multiple investigating how relies mechanisms, less attention combines programming execution. propose that, number, or (see Glossary), there exists subserving perception: system, tuned signals supported by range suggesting (regardless modality presentation format) process events, goal-directed routines. We review this type non-numerical dimensions items represented two distinct processes: cultural language-dependent encodes precisely cardinality elements; primitive variety species termed approximate (ANS), quantity (or numerosity) an approximate, non-symbolic manner [5.Dehaene Number Sense: How Mind Creates Mathematics, Revised Updated Edition. Oxford University Press, 2011Google Scholar,6.Burr D.C. et al.Psychophysical sense.Philos. 2017; 373: 20170045Crossref (39) Numerosity differs symbolic (digits) serial counting, it rapid, parallel, approximate. estimation errors scale proportionally physical numerosity, following Weber’s law, at least over limited [7.Ross J. Visual discrimination without counting.Perception. 32: 867-870Crossref (53) Scholar,8.Anobile G. al.Separate density.Psychol. 2014; 25: 265-270Crossref (99) opinion deals almost exclusively notion so proposed functional anatomical be those exact quantities. Humans not only but self-generated For example, mice learn press lever defined times reward, average responses variability increasing target repetitions (scalar variability) [9.Cavdaroglu B. Balci F. Mice count count-based decisions.Psychon. Bull. Rev. 2016; 23: 871-876Crossref (9) Bees distance nest feeder combining landmarks cues travelled distance, some them likely related internal variables estimates energy spent flight duration [10.Chittka L. Geiger K. Can honey-bees landmarks.Anim. Behav. 1995; 49: 159-164Crossref (150) Desert ants steps infer walked distance; if legs elongated shortened, they over- underestimate [11.Wittlinger M. al.The ant odometer: stepping stilts stumps.Science. 2006; 312: 1965-1967Crossref (289) Processing plays critical role frog mating behavior: female frogs select mates ‘chuck sounds’ make, turn affects male responses, must match exceed chucks competitors [12.Rose G.J. numerical abilities anurans correlates: insights neuroethological acoustic communication.Philos. 2018; 20160512Crossref (16) When asked reproduce series fast (without counting), mean linearly number. pattern results closely resembles version task participants verbally impulses rapid sequences flashes [13.Whalen al.Nonverbal counting humans: psychophysics representation.Psychol. 1999; 10: 130-137Crossref (413) Taken together, suggest actions characteristics function resemble associated stimuli. Further comes saccadic eye movements humans. At time movements, cloud dots become compressed [14.Binda P. al.Underestimation perceived saccades.Vis. Res. 2011; 51: 34-42Crossref (19) paralleling similar effects occur [15.Morrone M.C. al.Saccadic cause compression space.Nat. Neurosci. 2005; 8: 950-954Crossref (290) Scholar,16.Ross al.Compression before saccades.Nature. 1997; 386: 598-601Crossref (355) Fast saccades toward numerous arrays executed quickly (latencies short 190 ms), systems through pre-attentive operating feed-forward [17.Castaldi E. al.Fast eye-movements automatic direct.Proc. 2020; 287: 20201884PubMed Distortions manipulations numbers, additions subtractions when digits presented [18.Binda magnitude.PLoS One. 2012; 7e49587Crossref (13) substrates reasonably understood. seminal electrophysiological [19.Nieder A. Supramodal selectivity primate prefrontal posterior cortices.Proc. Natl. Acad. U. 109: 11860-11865Crossref (110) Scholar, 20.Nieder neuronal code number.Nat. 17: 366-382Crossref (162) 21.Nieder al.Temporal spatial enumeration cortex.Science. 313: 1431-1435Crossref (214) 22.Nieder al.Representation 2002; 297: 1708-1711Crossref (479) 23.Viswanathan Nieder Neuronal correlates “sense number” 2013; 110: 11187-11192Crossref (86) Scholar] shown single-cell activity cortices macaques specifically numerosities temporal (Figure 1A,B ). In humans, similarly recorded horizontal intraparietal sulcus (HIPS) fMRI habituation paradigm [24.Piazza al.Tuning curves sulcus.Neuron. 2004; 44: 547-555Abstract (695) Interestingly, monkeys, cells area 5 [located near ventral (VIP)] 1C [25.Sawamura H. al.Numerical 415: 918-922Crossref (196) Scholar]). Monkeys were trained perform arm movement (e.g., push) five times, then switch another turn) back initial movement. Around third selectively specific repetition routine, normalized number-action Figure 1C. Furthermore, transient pharmacological inactivation caused (while producing deficits), demonstrating causal [26.Sawamura al.Deficits selection based after monkeys.J. Neurophysiol. 2010; 104: 902-910Crossref (29) Given low involved (around 5), monkeys engaged rather than estimation, study nevertheless demonstrates clear action. There tuning medial lobe (MTL) neurosurgical patients performing calculation tasks [27.Kutter E.F. al.Single brain encode numbers.Neuron. 100: 753-761.e4Abstract (30) first operand subsequent delay period elicited hippocampus, parahippocampal cortex, entorhinal amygdala 1D) numbers. Most number-selective found highly interconnected frontal areas. It contains activated hand-grasping observation [28.Mukamel al.Single-neuron during actions.Curr. 20: 750-756Abstract (704) opening possibility could mediate interaction Imaging regions located nearby partial overlap areas (eye hand movements) [29.Simon O. al.Topographical layout hand, eye, calculation, language-related lobe.Neuron. 33: 475-487Abstract (599) Scholar,30.Hubbard E.M. al.Interactions cortex.Nat. 6: 435-448Crossref (837) used structural criteria define putative homologs macaque [31.Castaldi al.Attentional amplification codes independent quantities along dorsal stream.eLife. 2019; 8e45160Crossref (20) 32.Eger al.Spatially invariant coding subregions cortex.Cereb. Cortex. 2015; 1319-1329Crossref 33.Harvey B.M. al.Comparing quantity-processing macaques.Trends 21: 779-793Abstract (14) 34.Harvey al.Topographic 341: 1123-1126Crossref (248) 35.Knops al.Recruitment mental arithmetic.Science. 324: 1583-1585Crossref (261) 36.Harvey Dumoulin S.O. network topographic maps association Hum. 1: 0036Crossref (40) 37.Castaldi al.Mapping subcomponents cognition relation cortex.Neuroimage. 221: 117210Crossref (5) decoder distinguish leftward rightward homolog lateral (LIP) subtraction addition [35.Knops non-symbolic, nicely complementing showing influence arithmetic LIP, [32.Eger localized VIP motion tactile stimulation face multivoxel patterns evoked activity. Using phase-encoding mapping technique, six field IPS, (IPS1/2) candidate monkey anterior (IPS3/4/5) [38.Kastner al.A brief comparative cortex: novel hypothesis toolmaker.Neuropsychologia. 105: 123-134Crossref (37) These regions, superior/medial parts IPS superior gyrus, active visually [39.Castaldi operations cortex.bioRxiv. (Published online April 9, 2019. https://doi.org/10.1101/602599)Google 2A Population receptive-field [34.Harvey Scholar,36.Harvey revealed topographically (NCP1–3) selective numerosities. Like VIP, NCP1 region convergence 2B [33.Harvey correspond [40.Gentile al.Integration brain: FMRI study.J. 910-922Crossref (166) hosts control limb [41.Gamberini al.Structural connectivity properties lobule.Brain Struct. Funct. 225: 1349-1367Crossref (18) including Scholar,26.Sawamura recent map (area NCP3) recruited observed [42.Sawamura actions: study.Eur. August 3, 2020. https://doi.org/10.1111/ejn.14930)Crossref (2) Other multisensory homunculus central (PoCS), primary somatosensory extending into inferior PoCs sulcus. ‘out order’ VIP+ [responsive 2C); see [43.Huang R.S. Sereno M.I. Multisensory maps.Handb. Clin. Neurol. 151: 141-161Crossref (17) Upper lower limbs intermingled lobe. implicated visuomotor reaching grasping defensive feeding behaviors 2D). proximity supporting tapping will discuss detail. Overall, given integration guidance played behavioral Together maintenance working memory supramodal (visual, auditory, tactile) [44.Spitzer al.Parametric alpha- beta-band signatures memory.J. 34: 4293-4302Crossref (31) Scholar,45.Uluç I. memory.eNeuro. 7 (ENEURO.0090-19.2019)Crossref mediotemporal containing constitute substrate Although hypotheses largely speculative, favor interaction, size [46.Harvey representations object relationships reveal cortex.Proc. 112: 13525-13530Crossref (77) animal imaging find support leveraging technique (Box 1). Perceptual one simple, powerful, robust probe mechanisms: observing particular stimulus (adapter) usually misperceived, negative aftereffects. Adaptation aftereffects assumed reflect changes response functions feature [47.Barlow H.B. Hill R.M. Evidence explanation waterfall phenomenon figural after-effects.Nature. 1963; 200: 1345-1347Crossref (302) 48.Clifford C.W. al.Visual adaptation: neural, psychological computational aspects.Vis. 2007; 47: 3125-3131Crossref (222) 49.Thompson Burr aftereffects.Curr. 19: R11-R14Abstract (120) 50.Benucci al.Adaptation maintains population homeostasis 16: 724-729Crossref (92) hence underlying ‘if you adapt it, it’s there’ [51.Mollon J.D. After-effects brain.New 1874; 61: 4Google Scholar].Box 1Perceptual AdaptationPerceptual characterize typically involves phase participant passively observes adaptor judge test stimulus, comparing unadapted location. causes misperception aspects direction motion, biased opposite adaptor: few seconds downward waterfall, shifting gaze adjacent appear move upward [78.Addams An account peculiar optical phaenomenon seen having looked moving body.Lond. Edinb. Philos. Mag. 1834; 5: 1Google attributed imbalance relative activities vertical sustained motion. tap unbalanced, measurable illusory populations sensitive tested give organized, revealing direction, orientation, size, I), others [49.Thompson ubiquitous throughout all systems, representing form short-term plasticity, probably serving tool calibrate statistics [68.Solomon S.G. Kohn Moving beyond suppressive single neurons.Curr. 24: R1012-R1022Abstract (111) Scholar,69.Barlow synaptic mechanism aftereffects.in: Blakemore C.B. Vision: Coding Efficiency. Cambridge 1990: 363-375Google Scholar,79.Barlow Conditions versatile learning, Helmholtz’s unconscious inference, perception.Vis. 1990; 30: 1561-1571Crossref (103) 80.Webster M.A. coding.J. Vis. 11: 3Crossref 81.Webster adaptation.Annu. 547-567Crossref Ross [52.Burr sense number.Curr. 2008; 18: 425-428Abstract (347) demonstrated like attributes, susceptible [53.Butterworth speckles hen?.Curr. R388-R389Abstract After exposure dense dot array several seconds, was s