Tango fitts haptic interpersonal coordination

University of Connecticut OpenCommons@UConn Doctoral Dissertations University of Connecticut Graduate School 5-5-2017 Tango-Fitts: Haptic Interpersonal Coordination Lin Nie lin.edu Follow this and additional works at: https://opencommons.edu/dissertations Recommended Citation Nie, Lin, "Tango-Fitts: Haptic Interpersonal Coordination" (2017).edu/dissertations/1378 Tango-Fitts: Haptic Interpersonal Coordination Lin Nie, PhD University of Connecticut, [2017] Interpersonal coordination is sustained by meaningful informational coupling, whether optic, acoustic, haptic or some combination. Such information is specific to the guidance of perception-action in a given activity unfolding at the animal-environment scale. The social dance of Argentine Tango provides a rich interpersonal coordination setting to study such informational coupling with an emphasis, in particular, on haptic coupling. In three experiments, the classic Fitts task was modified to allow a continuous (not discrete) monitoring of error and a treatment of Index of Difficulty as an obtained (rather than imposed) value.

Three coordination challenges inspired by tango were investigated: direction of movement, type of perceptual support, and improvisation-like demands arising from unpredictable targets. As expected, dyads were influenced by the direction of movement but solo actors were not (Experiment 1 vs. Dyadic coupling that involved haptics (with or without vision) provided a better fit to Fitts’s law than coupling that was exclusively visual (Experiment 2). Varying target location and limiting the preview of it still preserves Fitts’s law (Experiment 3).

While solo actors were affected by whether they had a zero or one cycle preview of the target, dyads were not. Results were discussed with respect to the contrast between Claude Shannon’s construal of information—limited, syntactic, and inherently meaningless—and James J. Gibson’s construal of information—lawful, meaningful, and specific to organism-environment circumstances relevant to perception-action. Implications for the intersection of dance (particularly ensemble improvisation dance), human-computer interaction, and experimental psychology were also considered.

Tango-Fitts: Haptic Interpersonal Coordination Lin Nie B., Franklin and Marshall College, [2010] A Dissertation Submitted in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy at the University of Connecticut [2017] i Copyright by Lin Nie [2017] ii APPROVAL PAGE Doctor of Philosophy Dissertation Tango-Fitts: Haptic Interpersonal Coordination Presented by Lin Nie, B. Major Advisor ___________________________________________________________________ Claudia Carello, Ph. Associate Advisor ___________________________________________________________________ Adam Sheya, Ph. Associate Advisor ___________________________________________________________________ Bruce A.

Associate Advisor ___________________________________________________________________ Tehran J. University of Connecticut [2017] iii ACKNOWLEDGEMENTS The author acknowledges the following: My advisor Claudia Carello for her guidance throughout this work and for being a role model by perfectly blending artistic expressions with scientific exactitude; Michael Turvey, for empowering his students and holding us to the highest standard. Bruce Kay, Tehran Davis and Claire Michaels for asking inspiring questions; Adam Sheya for encouraging me to pursue this work in his Developmental Systems seminar; CESPA professors—Bill Mace, James Dixon, Till Frank, and Kerry Marsh—for growing me intellectually; Tony Chemero and the Psychology and the Dance Departments at my alma mater Franklin & Marshall College for exposing me to the coolest materials in those fields early on. Thanks to Lucas Dixon for prodding me and literally keeping me off the streets in the past year; all my friends on the east coast of the US, for their acceptance, wisdom and depth that lifted me through many of life’s traps in my late 20s: Kartikeya Rao, Yitong Wang, Nick Eppert, Alex Woodend, and Christine Putre; those CESPA friends for much-needed support during the CESPA years: Henry Harrison, Lily Shield, Vitor Profeta, Theo Rhodes, Annelies Rhodes, Treysi Terziyan, Stephanie Petrusz, Jason Gordon and Tosca Braun; and my parents overseas, Wang Fengzhi and Nie Li, for raising me right and loving and supporting me all my life.

The initial idea of this dissertation came from many car discussions with Dobri Dotov when traveling through the Rhodope Mountains in Bulgaria, summer of 2014. I thank Dobri for all the things that we created together. I save the burden of any shortcoming in this document for my own, and dedicate the thesis to dance and music. iv LIST OF TABLES Table 1.

Summary of Average Obtained ID by Each Participant as a Function of Movement Direction in Experiment 1………………………………………………………………………. Average Movement Times (MT), Error, and Fitts Line Fits (r ) for Each Solo Participant 2 as a Function of Movement Direction…………………………………………………………. Average ID Produced by Each Dyad as a Function of Perceptual Coupling and Movement Direction……………………………………………………………………………. Average Fitts Line Fits (r ) as a Function of Movement Direction and Coupling for Each 2 Dyad…………………………………………………………………………………………….

Average Movement Time (msec) for Each Dyad As a Function of Movement Direction and Coupling……………………………………………………………………………………. Average Error (mm) for Each Dyad as a Function of Movement Direction and Coupling………………………………………………………………………………………. Average Movement Time (msec), Error (mm), and Fitts line fits (r ) for All Solo and 2 Dyad Trials………………………………………………………………………………………. Average ID Produced By Every Solo (top) and Dyad (bottom) separated by Preview Condition……………………………………………………………………………………….

Average Fitts line fits (r ) as a Function of Preview Condition for Solos……………. Average Fitts line fits (r ) as a Function of Preview Condition for Dyads……………57 2 Table 11. Average Movement Time (MT in msec), Error (mm), Endpoint Variability (mm), and Fitts Line Fits (r ) for All Solo and Dyad Trials……………………………………………….59 2 v LIST OF FIGURES Figure 1. Schematic of the “open embrace” in the social dance of Argentine Tango………….

Schematic of the Solo Fitts Task in Experiment 1…………………………………. The relation between MT and obtained ID for seven solo participants (i., not averaged) as a function of movement direction…………………………………………………21 Figure 4. The relation between MT and error as a function of target amplitude in transverse (left) and medial (right) directions for seven solo participants (i. Schematic of the Dyad Fitts task in Experiment 2………………………………….

Data for the seven dyad (not averaged) showing the relation between MT and obtained ID………………………………………………………………………………………………. Data for the seven dyad (not averaged) showing the relation between MT and obtained ID for medial and transverse directions…………………………………………………………34 Figure 8. Data for the seven dyad (not averaged) showing the relation between MT and obtained ID for the three coupling conditions……………………………………………………………. Average dyad movement time as a function of distance and coupling for (left) medial and (right) transverse movement directions…………………………………………………….

Average dyad error over four distances with three coupling formats………………. Average dyad endpoint variability (SD of error) over four distances in three coupling formats…………………………………………………………………………………………. Schematic of the improvisation task in Experiment 3………………………………. Data for the seven individuals (not averaged).

(left) The relation between MT and obtained ID in zero preview. (right) The relation between MT and obtained ID in one preview. Average movement time as a function of distance and preview condition in solos…. Data for the six dyads (not averaged).

The relation between MT and obtained ID in (left) Zero preview, and (right) One preview……………………………………………………. The coordinate system used by the Fitts-Tango program………………………. Visualization of tap plots……………………………………………………………85 vii TABLE OF CONTENTS CHAPTER 1: INTRODUCTION .1 CHAPTER 2: TANGO BASICS .49 CHAPTER 6: GENERAL DISCUSSION .85 viii CHAPTER 1 INTRODUCTION When people engage in joint actions, the informational coupling can be optic, acoustic, haptic or some combination. As examples, navigating through a crowded airport is primarily visually guided, vocal ensembles are constrained by hearing the other singers, and furniture movers rely on touch transmitted through the object they hold.

The rigorous examination of joint action owes much to the pioneering work of R. Schmidt and colleagues (e., Schmidt, 2007; Schmidt, Carello, & Turvey, 1990; Schmidt & Richardson, 2008) in applying dynamical systems theory to processes of interpersonal synchrony. Research has revealed that stable organization and patterning of interpersonal coordination emerge from the information coupling between individuals and their environment (e., Fowler, Richardson, Marsh, & Shockley, 2008). While copious research has highlighted the sufficiency of vision in sustaining interpersonal coordination tasks (e., Fine & Amazeen, 2011; Richardson, Marsh, & Schmidt, 2005; Schmidt, Nie, Franco, & Richardson, 2014), information is available in rich acoustic and haptic as well as optic arrays (Turvey & Carello, 2011).

Very little attention has been paid to understanding how haptics, in particular, shapes ongoing interpersonal coordination. Haptic coupling provides the focus of the present dissertation. In particular, the research was guided by challenges inherent in the interpersonal coordination setting provided by dance and it exploited the methodology of a classical cyclical aimed­movement task owing to Fitts (1954). Dance provides a rich interpersonal coordination setting in which optic, acoustic, and haptic information all guide the dancers who see each other, hear the music, and are in physical contact with their partners.

Importantly for present purposes, we look at dances by relationships and 1 rules that can be implemented experimentally. In particular, the tango is a distinct dance practice that shares many qualities with how people act, perceive, and communicate in everyday life: Individuals come together to perform actions in coordination with other individuals and with the environment. What distinguishes tango from other partnering dances (e., salsa, ballroom)—and what makes it a unique inspiration for studying interpersonal coordination—is that the primary coupling is haptic. Specific details of tango’s movement grammar are provided in Chapter 2.

This dissertation used the coordination challenges of tango to constrain haptic collaboration in a generalized Fitts task to examine novel effects that arise in haptic interpersonal coordination. One advantage of the Fitts­Tango setting is that it allows haptic interaction research in a real environment. A dominant focus of haptic interaction research has been the utility of haptic feedback that allows individuals located remotely to collaborate in a shared virtual space. Within that focus, there is no direct haptic link even when haptics is critical to the tasks (e., co­manipulation of an object; Wang, Chellali & Cao, 2016).

This is unsurprising given that these experiments tend to stem from the fields of HCI (human­computer interaction), computer­aided design, telemedicine, and virtual environment gaming which historically have embraced a theoretical framework based on old­fashioned computer vision). When the haptic link is either simulated via a teleoperation system (Wall & Harwin, 2000) or mediated by machinery (Reed & Peshkin, 2008), it results in indirect coordination that is not comparable to the zero­lag mechanical feedback available in a partner dance. In using direct physical contact as the coupling medium, the present experiments examined how classical phenomena in a Fitts task were changed by the real physical demands (e., multi­dimensional movements, points of contact, availability of optic/haptic information or a 2 combination of both, predictability of target locations) observed in a partner dance such as tango. Generalized Fitts Task In a common Fitts task, an aiming movement is alternated continuously between two targets whose size and distance can be varied to manipulate the difficulty of the task.

Such a precision aiming task was appropriate for the present study because precision is one of the most prominent constraints in everyday coordination (Latash, 1993), such as navigating through animate and inanimate clutter so as to avoid collision. The choice of a generalized Fitts task also stems from the simple recognition that traversing a cluttered environment whether in a partnering dancing or by fundamental bipedal locomotion both require the cyclical “left­right­left­right” stepping motion or the “go­return­go­return” Fitts task motion. The cyclical Fitts task (also referred to as reciprocal aiming in the literature) does not require the pointer to come to pause on the target. In its simplest form, Fitts’s law characterizes influences on movement time, MT, owing to manipulation of target width, W, and movement amplitude, A, primarily dictated by inter­target distance: MT = a + b log2 (2A/W).

In particular, it is a prediction of movement time as a function of an index of difficulty, ID, captured by the ratio of A to W. As one of the most robust laws in biological motion, Fitts’s law (Fitts, 1954; Fitts & Peterson, 1964) has been shown to hold for different effectors, on many scales (Guiard, Beaudouin­Lafon, & Mottet, 1999), across differing pointing devices (Fitts, 1954), and to be applicable to hand and foot movements, in air and water, under a microscope, and along with other modifications (Hoffmann, 1981).