1
We're Going to Study the Mind!
James M. Brown and Harold H. Greene*
University of Georgia, *University of Detroit Mercy
Introduction
It all started for me (James) the spring of 1978, when, on the first day of my last required course for my undergraduate degree in psychology, Professor Naomi Weisstein introduced herself to her Sensation and Perception class and said âWeâre going to study the mind!â âReally,â I said to myself, ânow that sounds interestingâ. Over the following weeks I realized I was hooked. This was very cool stuff, real psychological science. I needed to know more. I asked Dr. Weisstein after class how I might get more involved in this mind stuff. She directed me to her lab where her graduate students trained me to be an observer in their ongoing experiments. As it turned out the experiments Amanda and Mary Williams (no relation) trained me to run in were metacontrast masking studies exploring the temporal and configural nature of the object superiority effect. In the beginning I was truly a naĂŻve observer in many ways. I wondered how what I was doing could possibly be of any help to them because I often felt I was simply guessing which of four target lines embedded in a figure I saw on each trial. Later during my first lab meeting, Naomi seemed thrilled with my results as much as I was amazed by them looking so systematic and similar to the other observerâs results. This was my first experience with and introduction to what Naomi later described to me as cognitive psychophysics. Towards the end of that semester I asked Naomi where I might go from here given how fascinated I had become with the study of visual perception. Without a momentâs hesitation she suggested I enroll as a graduate student and work with her at SUNY Buffalo. I was shocked and then so excited for the opportunity. From that first day of class and since, I have felt ever so lucky to have met this wonderful woman. As I have described to her beloved husband Jesse Lemisch, I was blessed to have crossed paths with this amazing woman, this ball of energy that was the âcomet Naomiâ. Somehow space and time aligned for me to have crossed the tail of the comet Naomi as it streaked across the visual-cognition sky and in doing so my life was changed forever. As we know though, when we see a comet, it is in the process of burning out. Thatâs why I think of myself as passing through the tail of the âcomet Naomiâ because soon after starting to work with her she began to suffer from what was later determined to be Chronic Fatigue Immune Dysfunction (CFIDS). Remarkably, as the comet Naomi was slowly physically burning out over so many years, her amazing heart and spirit continued to burn brightly. The beginning of her illness led to lab meetings in her living room while she laid on the couch, to her return to NYC with Jesse, and ultimately to her being bedridden for the next thirty plus years until her death in the spring of 2015. I often wonder where our field of vision science would be today had she remained healthy and been able to spend those years pursuing her study of the mind. It was her energy and enthusiasm for studying the mind that inspired me to study visual perception.
It is the seeming immediacy and effortlessness of visual perception that belies the amazing complexity of what underlies it. As we interact with our environment the stimulation received through our eyes informs us of the world around us. How this information is acquired and utilized by our brains resulting in our conscious experience is still a curious question. It certainly involves more than âsimplyâ taking in stimulation from the environment. One might consider our visual brains as âreaching outâ to engage with the world to be able to understand what is out there, analogous to reaching out with our hands to identify through touch something we are unable to see. In both instances the afferent stimulation via the senses is utilized within the context of directed brain activity to âgraspâ what we believe to be the nature of the world we perceive based on expectations and experiences. The expectations and experiences are a mix of variable short-term ones that precede and guide our sequential fixations as we explore the visual environment and stable long-term ones based on our evolutionary past and our knowledge of the world accumulated via lifetime experiences of acting upon and moving about in the world (e.g., Hoffman, 1996; OâRegan & NoĂ«, 2001; Purves et al., 2015). Trying to understand how this complex interplay between bottom-up stimulation and top-down experience and knowledge creates our perceptual experience underlies most of Naomi Weissteinâs research and illustrates her influence on our (Naomiâs academic genealogical son and grandsonâs) research.
A fundamental issue in vision research is trying to determine how we segregate figure from ground, locating and differentiating the objects from their backgrounds in our environment. Much has been learned since the early work of Gestalt theorists to the present day, yet how this is achieved remains a puzzle. Our understanding of the visual system has grown tremendously since those early days and the tools available to probe this mystery have grown in numbers and technical sophistication, yet we are still left wanting for a complete and clear understanding. In her last published work, Weisstein proposed a model of figure-ground perception based on antagonistic magnocellular (M) and parvocellular (P) pathway interactions in the visual system (Weisstein et al., 1992). The model emerged from research in her lab in the 1980âs and was based on what was known at that time about the spatio-temporal frequency and chromatic response properties of the M and P retino-geniculo-cortical pathways. Neural activity initiated in the retina and passed along the M and P pathways is utilized by subsequent brain areas (e.g., in the dorsal and ventral streams) in many different ways. Weissteinâs model foreshadowed a number of widely different discoveries related to the dorsal and ventral cortical streams, visual brain physiology and circuitry, and the dynamic, interactive nature of bottom-up and top-down processing. Here we revisit Weissteinâs model in light of these different discoveries while also extending it as a general framework for considering eye movement behavior when looking for a figure.
Figure-Ground Perception Based on Antagonistic Dorsal-M & Ventral-P Stream Interactions in The Visual System
At the time Weisstein offered her model there was growing debate about the extent to which the M and P pathways were parallel as well as how and where interactions between them might take place. The M pathway provides the dominant initial feedforward input to the dorsal stream and the P pathway the dominant initial feedforward input to the ventral stream (Haxby et al., 1991; Ungerleider & Haxby, 1994; Ungerleider & Mishkin, 1982). There is now sufficient evidence to argue that antagonistic interactions may be occurring both within the dorsal and ventral streams themselves and between structures in the dorsal and ventral streams (Appelbaum et al., 2006; Doniger et al., 2002; Donner et al., 2008; Wokke et al., 2014). Considering the mixing of M and P connections in V1 and in extrastriate areas found in macaques in both dorsal (Maunsell et al., 1990; Nassi et al., 2006) and ventral streams (Chen et al., 2007; Ferrera et al., 1992, 1994; Merigan & Maunsell, 1993; Nassi & Callaway, 2009; Nealey & Maunsell, 1994; Sincich & Horton, 2005), the neural architecture to afford the dynamic antagonistic interactions of M-dominant dorsal and P-dominant ventral activity Weisstein envisioned would appear to be there. Considering the mixed M and P geniculo-cortical inputs to the dorsal and ventral streams one could describe the inputs to the dorsal stream as dorsal-M and dorsal-P and the inputs to the ventral stream as ventral-P and ventral-M with the relatively dominant inputs being dorsal-M and ventral-P. To best represent the intention of Weissteinâs original M-dorsal/P-ventral distinction, we will use the terms dorsal-M and ventral-P when discussing her model and their antagonistic interactions related to figure-ground perception. How then does this relate to our perception of figure and ground?
The basic idea underlying Weissteinâs model is that activity in the ventral-P stream encodes figure/foreground information and activity in the dorsal-M stream encodes background information. The ventral-P streamâs figure-biased signal is therefore associated with a region being perceived as figure while the dorsal-M streamâs ground-biased signal is associated with a region being perceived as ground. In general, where a boundary separates two regions, which region is perceived as figure and which as ground is determined by the outcome of antagonism between dorsal-M and ventral-P signals both within each region and across the boundary between the regions (see Figure 1.1 for an example and Weisstein et al., 1992 for more details). The region with the relatively stronger ventral-P âfigure signalâ is ultimately perceived as figure while the region with the relatively stronger dorsal-M âground signalâ is perceived as ground (e.g., as represented by the relative circle sizes in Figure 1.1). We next summarize some of the original evidence supporting this idea followed by recent findings consistent with it.
Bottom-Up Influences
FIGURE 1.1 Figure-ground perception and the spatio-temporal frequency and chromatic response.
Figure-ground perception based on antagonistic dorsal-M & ventral-P stream interactions in the visual system. The perception of a region as figure involves top-down (dashed arrows) and bottom-up influences (circles). Antagonism across the boundary (black double arrow) separating two regions (vertical dashed line) occurs with ventral-P dominant figure and dorsal-M dominant ground signals generated from antagonistic dorsal-M & ventral-P stream interactions (gray arrows) within those regions.
One tactic for biasing processing towards dorsal-M vs. ventral-P streams is based on their different spatio-temporal frequency and chromatic sensitivities. The dorsal-M stream is relatively more sensitive to lower spatial and higher temporal frequencies while the ventral-P stream is relatively more sensitive to higher spatial and lower temporal frequencies (Derrington & Lennie, 1984; Kulikowski & Tolhurst, 1973; Legge, 1978; Tolhurst, 1975), and chromatic information (Livingstone & Hubel, 1987, 1988). There is overwhelming evidence indicating spatial frequency differences affect the perception of figure-ground (Klymenko & Weisstein, 1986; Klymenko et al., 1989) and depth (Brown & Weisstein, 1988b). In separate experiments, regions of ambiguous pictures (e.g., the Rubin faces/vase picture (Figure 1.2b), the Maltese Cross (Figure 1.2d), a bipartite field) were filled with sinewave gratings of different spatial frequencies ranging from 0.5 to 8 cpd at one-octave intervals yielding spatial frequency differences ranging from one to four octaves between regions of the ambiguous pictures. The influence of spatial frequency differences on figure-ground perception was measured by the percentage of time regions were perceived as figure. In general, higher spatial frequency filled regions were perceived predominantly as figure relative to lower spatial frequency filled regions, which were perceived as background. This bias to perceive higher spatial frequency filled regions as figure increased as the octave separation between regions increased. These findings also indicated that relative rather than absolute spatial frequency values determine whether a region will be perceived as figure or ground with the perception of figure always biased towards the higher spatial frequency region. The bias to perceive the higher spatial frequency filled regions as figure is accompanied by a strong impression of their being closer in depth (e.g., compare your experience when looking at Figure 1.2c vs 1.2d). This depth impression is so striking that observers consistently place lower spatial frequency filled regions closer in stereo depth to make them appear in the same depth plane as higher spatial frequency filled regions (Brown & Weisstein, 1988b).
FIGURE 1.2 Figure-ground contexts.
(a, b) Rubinâs faces-vase; (c, d) Maltese Cross; (b) and (c) are filled with sinewave gratings as used in experiments examining the effect of spatial frequency on figure-ground perception.
There is also substantial evidence indicating temporal frequency differences affect the perception of figure-ground and depth (Klymenko et al., 19...