Periods and Themes

 

UNC Graduate School: 1966-1970

Background.  I graduated with a B.A. degree in mathematics from The University of the South (Sewanee) in 1962, completed a M.A. in English Literature at The University of South Carolina in 1964, during which I married my wife, Catherine Findley, and then taught (English lit and composition courses) at Clinch Valley College - a two-year branch of the University of Virginia - from 1964 to 1966.  While at USC, our son, Ian, was born. In 1966, both Catherine and I began Ph.D. programs in English at The University of North Carolina in Chapel Hill.
In the first two years at UNC, I completed all of the course work and exams for the degree in English.  However, in fall, 1967, I took my first computer science course - a PL/1 programming course for liberal arts majors with emphasis on natural language processing.  It was taught by Sally Sedelow who held joint appointments in both the computer science and the English departments.  I was hooked!  I took a second course in the spring and began working as a research assistant for Sedelow in the summer of 1968.  Since I was through by then with English department requirements (except for the dissertation), I spent the next two years taking (and auditing) computer science and psychometric courses, where I first learned about some very powerful statistical analysis and modeling tools.  More important was the work I was doing for Sedelow's Automated Language Analysis research  project, where I became familiar with then state of the art techniques and programming tools for natural language computing. During that  period, I was also able to develop the software, analysis, and interpretive strategies used in the dissertation, which I completed in 1970.  Works from the Sedelow annual reports and the published version of the dissertation are listed below and can be downloaded from those references.

 Sally Y. Sedelow, Automated Language Analysis Project

Dissertation: Imagery and the Mind of Stephen Dedalus


PSU: 1970 - 1984

I went to Penn State in 1970, with  joint appointments in the English department and the Computation Center. For English, I taught conventional courses in literature, composition, as well as interdisciplinary courses, such as Science and Human Values.  For the Comp Center, I was a member of the applications staff, and my primary charge was to promote and support use of computers in the liberal arts, especially the humanities.  In 1972, I developed and began teaching a programming course for Liberal Arts students, similar to the one I took from Sedelow at UNC.  Over the next several years, I added two more technology and methods courses so that by 1977, I discontinued the Comp Center appointment and took a half-time position in the College of Liberal arts in order to teach this three-course sequence. 
During these Penn State years, I worked in some half-dozen humanities computing areas, listed below.  These included work in natural language processing techniques and technology, bibliography systems and practice (our Shakespeare bibliography project was one of the first  to support direct, computer driven typesetting), and computer-assisted analyses of literature as well as oral performances.  Another area I found especially challenging was literary critical theory.  I had become convinced that doing in-depth analyses and interpretations of literary works (and other types of texts) represented a fundamentally different mode of intellectual inquiry.  To explain this view, I wrote several articles describe\ing what was different about this way of working/thinking but also situating it within traditional intellectual traditions, including structuralist/formalist criticism and Marxist criticism.  The paper I never got around to writing would have attempted to explore the phenomenological implications for the critic, analogous to Wolfgan Iser's notion of "second readings," except on steroids!

Literary Analysis

Systems and Technology

Critical Theory and Methods

Folk Sermons Analysis

Bibliography


UNC: 1984 - 1995

During the latter part of my tenure at Penn State, I began to focus more on the computer or technical aspects of my work and less on the literature or humanities aspects. Consequently, I took a leave of absence in 1982 in order to spend the year as a visiting faculty member in computer science at The University of North Carolina.  At the end of that year, they offered me a continuing appointment, and I joined the department in January, 1984.  This began a period that was less focused on humanities research and more on hypertext systems, cognitive strategies and systems for knowledge construction (e.g., expository writing or system design), and collaboration technology and theory.  During this time, we hosted the first international conference on hypertext, Hypertext '87, and one of the early conferences on computer supported cooperative work, CSCW '92, for both of which I served as conference co-chair.
A key part of this work was the attempt to integrate cognitive theory with system design and to use the system as a data collection device.  To understand writers' cognitive strategies, we developed a number of tools including cognitive grammars to analyze system-recorded protocols.  This approach was used first with respect to technical and professional writing.  It was then extended to groups, where we added a social or anthropological perspective to the cognitive in order to understand how groups (sometimes) develop seamless, integrated  conceptual structures - i.e., produced from a type of "collective intelligence." A hallmark of this work was to base the user's functional model of the system on an underlying mental model of the task - for writing systems, a cognitive theory of writing based on different cogtnitive modes, and, for collaboration systems, a cognitive/social theory of collaboration with emphasis on the artifacts developed to accomplish the task.

 Hypertext Writing Systems, Theory, and Research

Writing Instruction

Collaboration: Systems and Theory

Query Reformulation

Text Systems

Data Analysis Environment


UNC: 1995 - 2011

Background.  In 1989, Tim Berners-Lee wrote a withepaper describing an Internet-based hypertext system, which he called the WorldWideWeb, to be developed at CERN.  Its primary purpose was for sharing scientific papers, particularly among high-energy physicists.  He then wrote a simple proof-of-concept implementation of his ideas in early 1990.  His work quickly attracted attention and users, even though it could only be accessed at that time through a command line client. In 1993, the first graphical, point and click browser appeared.  Called Mosaic - later to be named Netscape - it dramatically increased the speed and convenience of the Web. And, as they say, the rest is history.
I first became interested in the Web in the 1994-95 period.  Unlike other hypertext systems, which were most often single computer applications, the Web was an architecture - that is, a set of protocols that anyone could implement and deploy on the Internet.  The three primary protocols or standards were the URL, which describes the address of an object on the Internet; HTTP, which describes a transfer protocol for Web clients (browsers) to communicate with Web servers; and HTML, a standard for basic formatting of documents that could be implemented in diverse browsers.  Another of the main reasons the Web succeeded so spectacularly versus other hypertext systems is that the Web simply ignored several particularly thorny problems that the hypertext community had viewed as fundamental.  Among these were inherent support for authorship and protocols for insuring that links were viable (no "dangling" or "broken"  links). These assumptions or requirements were simply left out of the architecture of the Web, to be provided through other, undefined means or not at all.
I became convinced by 1995 that conventional hypertext had been superseded and that if it were to retain any relevance, it would have to redefine itself within the context of the Web and implement its ideas as extensions of the Web architecture. Thus, I believed that we were living  through a paradigm shift.  I also believed that a large proportion of computing would shift from stand-alone systems to functional systems evoked and accessed through the Internet, with the Web serving as or providing the user interface. As a result, I shifted my activities from designing and building stand-alone hypertext systems to designing and building Web-based systems, and from writing and publishing in hard-copy form to writing and publishing on the Web.

In the interval between 1995 to 2000, I developed a three-course sequence of computer science courses concerned with such systems.  Although they evolved, their basic form and respective purposes remained the same.  The first course was concerned with building simple client-server applications using low-level tools such as Perl and CGI, soon replaced by JavaScript and PhP.  Both configurations usually fronted a database system such as MySQL. These configurations supported simple applications often involving rudimentary database queries with little programmed function.  The second course was intended for developing applications in which more extensive processing was required on the server side. At about the same time the Web took off, the Java programming language appeared. It was a powerful, object-oriented language with extensive support for Internet-based applications. Consequently, this second course was based on Java and a Java server.  Both of the first two courses presumed applications and/or services for which a single server would be sufficient.  For larger, enterprise applications, particularly those with high traffic volumes requiring multiple servers, the third course used the IBM WebSphere platform (also Java based) and its DB2 database.  Thus, it dealt with more sophisticated system designs and used more complex, more powerful "industrial strength" tools and middleware. 

To support my teaching in this area for the next 15 years or so, I developed some 80 online lessons that explained basic concepts but also illustrated implementation details through extensive code samples, with commentaries.  Unfortunately, most of those lessons were lost when they were deleted by a Comp Sci staff member, without notice or giving me the chance to move them elsewhere. The samples included below are a remnant of some early lessons I had saved in personal disk space. They are dated, some of the images are missing, and some of the links broken, but they give a taste of this work.

Web-Based Systems and Instructionn

Automatic Generation of Enterprise Systems


Retirement: 2011 - Present

I have been retired since 2011.  During that time, most of my time and energy, as well as my wife's, Catherine's, has been spent on our farm, ChicoryLane, located in rural central Pennsylvania.  Never a very productive agricultural plot, it is a wonderfully diverse habitat for native plants, wildlife, insects, and quiet, pleasing natural landscapes - both small and close as well as distant. Our understanding as well as feelings for the place have deepened during these years. In trying to understand them, we have settled on the notion, a sense of place. That is, we are trying to consciously develop an evolving, multifaceted understanding and appreciation for one simple, undeveloped place. A catalyst as well a tool for recording information and thoughts about the place is a Web site and database we have created.  (It is also an outlet for my programming urges.) It includes slideshows of the place and events held here.  A blog records musings about ChicoryLane - our sense of this place. A database catalogs over 300 native plants and a Google map lets us see not just the farm as a whole but also the locations of specific plant. We invite you to browse ChicoryLane by clicking on the link, below. (To gain full access to the database, please register and create a login.)

ChicoryLane Farm: Web Site, Database, and GPS System


Final Thoughts

Over the years, I have been fortunate to have a number of outstanding collaborators. Many are included as co-authors in the bibliography.  Others, most notably students, are not named but are remembered, and I thank them.  I especially wish to thank Geoffrey Rockwell; it was his idea to pull together this archive, and his encouragement has made it happen.  Finally, I thank my wife, Catherine, for her collaboration, contributions, and comments for more than fifty years.