Reading and Searching: Tools and Skills

François Lachance



How can educators and students take fuller advantage of the processing possibilities offered by full-text databases. What do you do for example with a single e-text such as a play by Shakespeare, let alone the collection of commentaries on the Divine Comedy offered by the Dartmouth Dante Database?

What kinds of tools can users apply? What kind of skills do they need to apply them well?

TACTWEB offers an interesting case. Such text analysis software enhances a user's willingness to conduct re-iterative and successive searches on the same body of material. In a computer-mediated communication environment, sustained interest in a given body of material can be further encouraged by exploiting the possibility of sharing results with other users.

Social interactions can be harnessed to motivate student learning. The educator can foster unique but structurally similar activities and events. The educator can help learners elicit the questions necessary for successful re-reading. The educator can also can model meta-cognitive behaviour and demonstrate how to ask questions about the results of a search in order to continue exploring. Indeed such modeling becomes internalized by the learning self and demonstratable to others.

An Example Module

If you prefer working inductively immediately, you may wish to explore an example exercise and discover how best to adopt it to your needs. The example is based on TACT (Text Analysis Computing Tools) in its Web version. TACT allows users to query a text or a full-text database. Once they have mastered the query syntax, users can quickly acquire information as word or phrase counts, distribution frequencies, collocations as well as view such information in a variety of formats ranging from keyword in context (KWIC) to simple graphs. Although I use TACTWeb as one way to foster a taste for re-reading in my students, many of the basic principles can readily be applied using the search and replace functions of wordprocessing software with or without a spreadsheet program to tabulate results. It is also important to take advantage of the learning possibilities offered by the software's help menus and online manuals.

TACTWeb is introduced after students have demonstrated skill in using a WWW search engine that permits Boolean searches. They have been also been introduced to the use of wild cards [(?), (*)] in file management. As well, a tour of Dartmouth Dante Database Project precedes the introduction of TACTWeb.

The work with search strings is designed to make students sensitive to variants in spelling, make them aware that word forms also come with prefixes and suffixes, and to provide guidance on how to use word combinations to find interesting collocations. With the Dante Database telnet://library.Dartmouth.EDU, I demonstrate a search for "man" followed by one for "woman" and for "child". The sheer difference in the number of occurrences provides a dramatic instance of the kinds of snapshots available using such a query tool. I then find the intersection of two of the terms and then all three. I also point out that the Dartmouth Dante Database covers commentaries in Italian and English and so introduce the question of multilingual searching. This is also raised in the process of checking chronology: I invite the students to consider when the term "woman" starts occuring in the commentaries. I use the Dante Database because, with its full text of the Divine Comedy and some 60 commentaries covering over 600 years of scholarship, its complexity makes the challenge of working on a single text via a WWW interface seem a bit easier.

For the TACTWeb assignment, students work through the tutorial on their own. The exercise is designed to contain two collaborative components. Students are to report on their group decision-making. In the first component they are to discuss the possible ways of citing the text they are exploring. When the text is a play, I ask them to choose some combination of citing (act, line, page, act/scene, speaker, scene) and to explain the basis of their choice. At this stage they are co-explorers of the interface and reporting possibilities. The second collaborative component is the last and most important element of the exercise and is weighted accordingly. Students are to show how the dialogue between the readers generates questions about what terms to use in the next search step. In short, they are expected to corroborate their interpretations with their peers.

The exercise forces re-reading. Students must consult and often re-read the tutorial material in order to master any given aspect of the suite of tools. This mastery is guided by specific assignment questions. The invitation to re-reading is ever present. I provide links to the help menus at the top of the assignment document. Furthermore, re-reading of both results obtained and steps taken is a must for a successful tracing out of the successive steps of a sustained inquiry. This is complex mix of skills. Details never standalone. A single view is never sufficient. The lone query is but a beginning.

After such an introduction, students might be invited to prepare (mark up) a text or texts of their own choosing to conduct their own analyses.

General Principles

A bit about the environment...

Students and teachers come to online learning with a set of habits and a body of folklore. These givens affect which pedagogical activities will seem novel and those which will seem familiar, which will need greater support and those which can draw upon group and peer experience.

Teles and Druxy (1991) cited in Paulson (1995) offer a listing of the most often used resources in networked environments. They are in descending order: email, computer conferencing, databases. It is possible to extrapolate from this particular study and surmise that the decreasing frequency of usuage correlates to an increasing possibility for disorientation. Email is asynchronous. Messages can be read and composed in a leisurely fashion. As well, email usually represents an interaction between self and a known interlocutor. Conferencing, be it via a MOO or Web Chat, exposes the self to the possibility of meeting unknown interlocutors. As well, conferencing can demand a certain level of skill to keep up with the pace of participation. However, participants in a conference can send and receive cues to guide behaviour and aid mastery of the interface. The interaction with databases can be construed as the least communicative of the activities listed by Teles and Druxy.

Adept multi-taskers will combine all three types of activities in any given session. Such ease demonstrates an ability to navigate shells or windows. It may also display a user's understanding that computers can and do run applications concurrently. Moving about is only one useful skill. Setting markers is another. Archivally-oriented users will be able to file email systematically, log and file conference sessions, save search strings and results.

These interactions are not mere technicity. They provide a basis for the acquisiton of content or the developement of critical acuity. The nitty gritty of electronic note taking, filing and retrieval invites at every step questions about the value and organization of information collected and presented. As in any environment, considerations of time and storage space do impinge upon the user's interactions. There is also the communicative dimension to consider: skills can be transfered and feedback rendered.

The networked environment offers many opportunities for presentations, actions and interactions. In pedagogical terms, the environment is very much like a laboratory setting where learners can witness demonstrations, can attempt experiments, can report back findings and can ask follow-up questions. The environment is conducive to a "what if" frame of mind. When a certain degree of comfort is established, learners can and do develop their own exercises and tasks.

An abstract look at activity

In networked environments, the mutability lends itself well to the creation of metacognitive moments. Users can easily ask, themselves or others, how to do something and what they want to do. As Bates describes the reformation of user objectives:

searchers are not just modifying the search terms used in order to get a better match for a single query ... the query itself is continually shifting (cited in Carmel, Crawford and Chen, 1992)

Of course, such a phenomena is epitomized in the lure of surfing the hypertextual universes of the World Wide Web. As Widemouth explains:

Because hypertext databases support traversal of the network of nodes so directly, they are also more likely to be used for browsing, while traditional databases are more often seen as a tool for identifying items that meet particular pre-specified criteria.

Full-text databases likewise support both searching and browsing. But the difference between browsing and searching is not clear cut. It is well worth considering the design of pedagogical exercises to incorporate objective-driven searches as well as reports on the serendipitous find.

The easy shift between searching and browsing has led to typologies of behaviours such as the triple categorization of search-oriented browse, review-browse, scan-browse (Carmel, Crawford, Chen, 1992). The key for pedagogical situations is review. This can take the form of peer exchange or journal entries that record the steps of a research process. There is, however, also a place for browsing without explicit goals. One user, in this fashion, found interest in launching searches based on misspellings. Play becomes productive. It stimulates the "what if" frame of mind.

As Carmel, Crawford and Chen point out, it is also possible to theorize a task according to the GOMS model developed by Card, Moran and Newell (1983):

knowledge states
elementary perceptual, motor or cognitive acts whose execution is necessary to change the user's mental state or to affect the task environment
a condidtional sequences of operators and goals; a learned procedure
Selection Rules
conditions under which methods are applied
Carmel, Crawford and Chen (1992) summarize how methods (previously acquired skill) and selection rules (knowledge of the task environment) affect action (operators):
Goals are achieved by applying selection rules to use the appropriate methods. (874)

In a learning situation, the selection rules may uncover an insufficiency in the methods and a subordinate goal may be introduced (e.g. acquiring knowledge of the proper syntax to conduct a search). The computer interface may offer multiple methods for accomplishing the same goal. Pedagogical exercices can draw upon this inherent forking. In their design, they can alternate between asking learners to explain how something is done and asking learners to provide the results of their activity. The alternation can occur over both group and solo components of an exercise. There is a wonderful metacognitive moment in group exercises where the doers and the readers are in dialogue: when they are consulting the manual and must decide to try an action or to read more or to repeat with variation.

Another look at reiteration ...

As noted above, a user's personal cognitive developement can be harnessed to the communicative capacities of the medium. As well, the resouces and tools available in networked environments permit users to modify and explore search stategies and tactics. The variety of experiences can not but enrich the exchanges likely to occur between users. It may be obvious but it bears restating: the possibility of repetition, the opportunity to replicate another's results, the minute methodicalness of computer processing, are what permit the variety of human-computer interactions and the attendent human-human exchanges.

In their ten steps outlining a basic search strategy, Ackerman and Hartman urge the reader to "Keep in mind that you can always modify your search at a later time." This invocation of the possibility of choice is echoed through out the ten steps. They encourage would be searchers, for example, to keep the selection of keywords and synonyms open. Furthermore Ackerman and Hartman stress the importance of revisiting steps.

I want to elaborate on the skills required to successfully modify a search expression. They would include:

I offer them in this order for a reason. A module can be built to take advantage of what we might call the "brag effect." The recording (simple cut and paste) of search strings makes them highly swappable by email. The results are replicable and verifiable by other users. A list of search strings can make the thinking process explicit to oneself and to one's collaborators. This becomes a very valuable source of ideas for search strings to mix and match when faced with conducting further searches on a set of results. From mere exchange, users can move to assessment. Note that assessment of precision involves knowing what a given search string missed and entails a move from comparing strings to reading results to re-reading sets of strings and results.

As basic as a list

Much of the skill involved in searching and re-reading involves the maintenance of lists. It is the work of mnemonics. It is also the work of parsing.

Working with lists is a step towards the generation of complete and grammatical sentences and even sustained prose. The notion of list introduces students to a flavour of modularity. They come to appreciate the building block nature of their interactions.

Once students discover that lists mutate and their private stock of search strings and results -- their own lists -- are exchangeable, they are well on their way to becoming seasoned communicators, researchers and project managers.


Owen reminds us that "good searching skills can assist students in becoming efficient, competent and life long users of information." He also suggests that one avenue to achieving this competency rests in exploration of the World Wide Web. This reminder is accompanied by great wariness of the metaphor of surfing. Owen prefers the French concept of "butiner," that is, to travel as a bee does gathering nectar and pollen from the flowers it encounters. For Owen, and for myself, "[e]ffective use of the Web in an education setting is very much a collaboration and communal effort." (Owen, 1996) . If hive mentalities do not appeal and the hum of the worker bees is alarming, you might want to consider Bates's invocation of berry picking. (Bates, 1989). You might even want to set a little exercise involving the chronology of competing metaphors. After all, a search engine awaits somewhere online.


Ackerman, E and K. Hartman. Finding and Evaluating Information on the World Wide Web (1995).

Bates, M.J. "The design of browsing and berry picking techniques for the on-line search interface". Online Review 13:5, 407-431 (1989)

Card, S.K., T.P. Moran and A. Newell. The Psychology of Human-Computer Interaction. (Hillsdale, New Jersey: Lawrence Erlbaum Associates, 1983)

Carmel, E., S. Crawford and H. Chen. "Browsing in hypertext: A cognitive study" in IEEE Transactions on Systems, Man, and Cybernetics. 22:5 (September/October 1992), 865-83.

Owen, D. Online Search Strategies and the World Wide Web (1996)

Paulson, M.F. The Online Reprot on Pedagogical Techniques for Computer-Mediated Communication (Oslo, 1995)

Teles, L. and N. Druxy. The Networked Classroom: An Assessment of the Southern Interior Telecommunicatons Project (SITP) (Vancouver: Faculty of Education, Simon Fraser University, 1991)

Wildemuth, B. Defining Search Success: Evaluation of Searcher Performance in Digital Libraries

Useful URLs

An example assignment:


The TACTWeb Home Page

A site in Cologne hosted by L. Mainka has all of Shakespeare as well as Joyce's Ulysses, Bronte novels and Bulfinch's Mythology

Lloyd Benson's collection of 19th century historical documents including among other texts Frederick Douglas, What to the Slave is the 4th of July? (1852)

Dartmouth Dante Database

At the prompt type "connect dante" and press enter.

R. Wallach and L. McCann provide in their article Weaving the Web into Course Integrated Instruction offer a very insightful experience of preparing a teaching module using the Dartmouth Dante Database as well as ARTFL.

An Off-line Forum

Information Doors -- Where Information Search and Hypertext Link

May 30th 2000
San Antonio, Texas, USA

A workshop organized by Einat Amitay (Macquarie University & CSIRO) and held in conjunction with the ACM Hypertext conference

Einat Amitay writes:

The task of reading from a screen is not a trivial one, nor is the task of navigating between online texts.

François Lachance
March, 2000