The Reading Matrix
Vol3, No.3, November 2003

Abstract

Reading electronically encompasses several major puzzles: what is reading, can it be taught, if so, how can it be taught, and how can electronic media such as the Internet hinder or enhance learning to read and teaching reading to others? Electronic enhancements present excellent opportunities to practice reading and a variety of these are examined in detail, including vocabulary and grammar acquisition activities, such as concordancing; skimming and scanning; extensive reading in WebQuests; and chaos management in audio, video, and text chat. Electronic media, which are categorized into three main types--text repositories, electronically mediated texts, and authentic computer-mediated communications (CMC)--present severe challenges to learners, while providing a rich storehouse of authentic reading matter. Several solutions offer help: Webpage design that takes learners' needs into account, appropriate teacher education to exploit electronic media to its fullest, and an emphasis on inquiry thinking or constructivism.

The Reading Puzzle

Some general agreement about what constitutes "reading" exists in the TESOL community. Based on extensive L1 research (especially in English; see Goodman, 1967, and Smith, 1971 ), most L2 researchers concur that reading involves a complex of skills and higher order cognitive processes. (See Chun & Plass, 1997 , for a summary of the L1 and L2 research) As an indication of its complexity and elusiveness, reading comprehension is often described in metaphors:

• bottom-up  - decoding through the automatic recognition of letters, vocabulary, and grammatical structures,
• top-down - understanding more global meanings, based on the reader's reasoning skills and prior knowledge, both of the text andthe world in general, including socio-culturally embedded meanings, and
• interactive - the learner actively uses an admixture of decoding skills, reasoning, and prior knowledge to interact with the text, building an "analog" mental counterpart to the meanings embedded in textual representations. ( Chun & Plass, 1997 , pp. 61-63)

None of these processes is linear or sequential; they all involve both short- and long-term memory functions; and one process or another may come more actively to the fore mentally to assist the others in the courseof processing a text. (For excellent summaries of the reading process, see Chun & Plass, 1997 ; Purcell-Gates, 1997 ; and Esky, 2002 .)

Less general agreement exists on the effectiveness of various instructional approaches in teaching both decoding skills and the higher cognitive and metacognitive functions of reading. Susser & Robb (1990) suggest that reading comprehension skills simply "do not exist" (p. 2). Krashen, in a variety of fora, has posited that the teaching of decoding skills does nothing to teach comprehension, and that post-reading activities, such as writing summaries, do nothing to improve reading speed or comprehension (see Mason & Krashen, 2002 , for the latest expression of this idea). For teachers who do use vocabulary and grammar practiceas part of "reading" instruction, there is no clear agreement on best practices. (See Nation, 1994, for most of the gamut of vocabulary teaching strategies.) As with vocabulary acquisition, no agreement exists on a single best approach to grammar development or even whether to begin with simpler structures first or the more complex, particularly at advanced learner levels (see Muckley, 1962 ). (For a balanced summary of these varied positions on teaching reading as a process, with a focus on comprehension, see Purcell-Gates, 1997 .)

For the higher level functions of reading, there is somewhat more concurrence on good practices: skimming and scanning activities,and pre-reading prompts (such as a picture or video) and questions to help with schema building, graphic organizers to help students successfully process information, and post-reading questions and activities or tasks such as essay writing, to help consolidate understanding. All of these purport to simulate actual processes used by experienced readers. As with most higher order language activities, teacher intervention during reading may disrupt the process itself (notice how students may become tongue-tiedwhen their grammar or pronunciation is corrected while speaking). Since reading does not "generate a product" ( Esky, 2002 ,p. 8), teachers usually must settle for related, post-reading products,e.g., answering comprehension questions, discussing the reading, or writing an essay using the material that has been read. We might add that unlike oral communication problems, breakdowns in reading comprehension are not immediately recognizable, so interventions to correct misunderstandings may take place long after the reader has set aside the text. The problem with teaching reading, as Esky puts it, is that "No one can teach someone else to read: The process is largely invisible and thus cannot be demonstrated,and it mainly occurs at the subconscious level and thus cannot be explained in any way that a reader could make conscious use of" ( Esky,2002 ,pp. 8-9).  Certainly, the evidence of studies in Fiji and Singapore (extensive, in-class, silent reading of high interest texts over 12-36 months by several thousand students; see Elley, 1991 , described in Nation, 1997 ) reveals that "bookflood" students not only did better on examinations based on vocabulary knowledge and grammar, but also wrote better. Perhaps even more importantly,they "enjoyed reading" ( Nation, 1997 ,p. 2).

Overwhelming evidence indicates that reading is the best way to learn to read, leaving in doubt the role of teachers in the classroom. However, Esky notes that besides motivating students to read more, teachers can teach "productive reading strategies," and refers to Anderson's excellent list of strategies, such as activating students' background knowledge, assisting them in acquiring more vocabulary, helping them select appropriate reading materials, checking for reading strategy use, and so on ( Anderson, 1999 , p. 6, cited in Esky, 2002 , p. 9). One suspects that much of this effort with ESOL (English for Speakers of Other Languages) students may be an attempt to counteract previous instruction in poor reading habits, such as reading word-by-word, looking up every word before proceeding, or subvocalizing the text or "reading aloud" (an oxymoron). Although one may suspect that reading is a virtually unteachable metacognitive process that develops holistically in the presence of readable texts, teachers need to know which aspects of the process are teachable, how teachable they are, and what the appropriate teachable moment might be for any given aspect with any particular student. I share Susser & Robb's view (1990), that language instruction is an essential part of ESOL (English for Speakersof Other Languages), but that many of the activities we perform in the reading class are not "a reading lesson in the strict sense" (emphasis by the authors,p. 1). I refer to the multiple definitions of reading, the complex cognitive,metacognitive, and emotional processes of reading, the difficulties in intervening in--and even teaching--reading, and the many possible approaches to reading instruction as "the reading puzzle."

By the early 1980s, even as the process approach to teaching reading was evolving, technology was making itself increasingly relevant to the solution, or at least to a better definition of the reading puzzle, first with computer-based programs to practice various aspects of decoding skills and cognitive processes, and almost immediately thereafter, as cyber-communication expanded, with computer-mediated communicative(CMC) approaches via the Internet and local intranets. As is generally the case with computer-assisted language learning (CALL), early electronically-enhanced approaches to teaching reading replicated most of the history of paper-based reading instruction. (For a general history of CALL development, see Hanson-Smith, 2002 .) The following section of this paper will look at some electronic approaches to teaching reading, focusing on the assumptions made about decoding skills and cognitive development. Because there are hundreds of pieces of excellent software and thousands of intriguing Websites intended to teach reading comprehension and decoding skills, I can give only a rough idea, tthrough some representative examples, of the kinds of technologically enhanced activities available. Since reading teachers have such widely varying needs and presuppositions, I have not pre-judged the various electronic activities, but rather presented a sampling of those that seem to use electronic media more effectively than activities that could be done on paper.

Electronic Access to the Reading Puzzle

Early computer software, which included all kinds of grammar tutorials and drills, vocabulary games, and tests, also sought to teach reading in ways that provided some major improvements over paper-based instruction. For example, for vocabulary practice, a number of free or nearly free programs produce crossword puzzles, hangman, bingo, concentration, and other types of games. The teacher enters a word list and the computer does the work of creating the game or challenge, replete with attractive sound effects and animations. An enormous advantage over paper versions is instant correction and user-tailored help. Vocabulary games attempt to give additional language exposure and practice in order to build the automatic recognition needed to read fluently. (Many of the older, mostly free games collected by TESOL's CALL Interest Section can be found at La Trobe University's CELIA site.)

Computers also excel at creating and managing tutorial and drill grammar activities, again with the advantage of providing instant corrections or hints as the student requires, and additionally, the possibility of preselecting appropriate levels or grammatical items to work on. Multimedia enhancements are intended to motivate and ensure a more memorable experience. Activities involve sentence combining by drag-and-dropping words or phrases, grammar-based games (similar to those mentioned for vocabulary acquisition), animated grammar where sentence parts move around, and others. Hundreds of sites on the Internet provide access to multimedia games and activities,traditional grammar drill-and-grill, searchable grammar reference books,and live tutors. Links can be found easily at Linguistic Funland ( Pfaff-Harris, 2002 ), and the Ohio University Language Lab ( Ohio ESL ). A few sites worth a try are Grammar for English Language Learners at OhioESL (also has links to grammar resources for teachers), HyperGrammarfor more advanced students ( Megginson, 1994-96 ),and Animated Grammar Tutorials at ESL Blue(s) for low level learners ( Pritchard, 1999-2002 ). RealEnglish allows teachers and students to download short videos and extensive accompanying grammar exercises, as well as vocabulary and culture lessons.

Cloze passages are often used to help students with the "psycholinguistic guessing game" ( Goodman, 1967 ) of reading by encouraging them to predict words they don't know, rather than stopping to look them up in a dictionary. While teachers can laboriously create cloze passages from authentic texts, by whiting out words by hand, many software packages, such as Wida's Gapmaster (in The Authoring Suite ), or the shareware SuperCloze ( Stevens & Millmore, 1990-96 ), allow the user to import text files and create word, letter, and even whole passage, cloze texts on screen. Additionally, the computer can insert the letters or words selected by the user into the text thus demonstrating to students that parts of words are of general help in decoding meanings), and offer hints or a peek at the target text when users reach frustration levels. Cloze passages can also be constructed to delete grammatical markers or punctuation. Jumbled sentences and jumbled paragraphs are another feature of computerized text reconstruction software. The unscrambling activity is intended to help students learn to recognize the structural markers that organize longer texts.

Computer software is ideal for its ability to present texts as timed or paced readings (the latter with scrolling text): The teacher or student can set a target reading time, presumably increasing it with each text (see Figure 1 for an example).The idea is to encourage students to eventually read fast enough to obtain global meanings (about 350 words per minute). Timed or paced reading can increase reading speed, or at least give students the metacognitive concept that they can and should be reading faster. In the classroom, the procedure is laborious: the instructor writes a series of times on the board and erases each as the seconds tick off; students lookup and record the last unerased time left on the board before turning to comprehension questions. After everyone completes the reading, the questions are answered and discussed. The activity takes up a great deal of class time, but has to be performed several times a week to produce any improvement in speed. Moving this type of activity to the computer lab (or home computer) is a huge bonus. It can become an almost totally autonomous student task, and some programs can also keep track of student progress and generate a report for the teacher as a double-check.

Skimming (picking up main ideas) and scanning (seeking specific details) are other activities generally thought to encourage students to read more rapidly and purposefully. These are available in many software packages, either with pre-set texts and questions or in a teacher utility that allows the instructor to load in texts and questions. Skimming is somewhat more difficult to design, since the questions usually involve inferencing. A good example of teacher-created products on the Internet is found in the Wild Children activities ( Mason, 1997 )generated with Hot Potatoes software (see Figure 1 ). Here skimming is combined with timed reading,consisting of a scrollable text passage with a countdown clock and a series of questions in a separate frame which can be answered as the student reads. Wild Children is one of dozens of teacher-prepared texts, organized by course level, at the University of Victoria's English Language Centre Study Zone. Activities also include other pre- and post-reading tasks, among them MCQvocabulary questions, interactive sentence combining using point-and-click to arrange words, comprehension questions, and composition topics. The quality of the questions prepared by teachers is one of the best features of the site.

Figure 1. Wild Children ( Mason, 1997 ) exemplifies an Internet-based skimmingexercise for advanced intermediate students (constructed with Hot Potatoes ). (English Language Centre Study Zone, University of Victoria, BC, Canada: http://web2.uvcs.uvic.ca/elc/studyzone/ ).

 Another kind of reading activity, one that is almost impossible to do with paper, is concordancing. The computer software searches for all examples of a word or phrase or word ending and then arranges all examples found as a list on the screen, usually with the found items in the middle of the screen so that a sentence context of about 10 words surrounds each one. A free piece of software, Conc , provides an example (see Figure 2 ):

Figure 2. Conc showing a concordance based on "had" in Lewis Carroll's Alice in Wonderland. Highlighting a line in the lower window highlights the corresponding target individual item in the upper window.

Grammar Safari ( Mills & Salzman, 2002 ) is an online activity very similar to concordancing, but on a less automated level.It uses the Find function on any Internet browser to search an electronic text. When the target word or morpheme is found, students copy the sentence by hand. (Copying may afford bonus practice in sentence structure and handwriting.) Using the lists of words in context, whether generated by the computer or written by hand, students can devise rules, or check their own usage of a structure in their essays against the authentic examples. Concordancing is an inductive approach to understanding vocabulary and grammar for the intermediate-advanced student, and vast corpora (as well as numerous how-to articles) are to be found both in software packages, such as MonoConc , and in digital text repositories on the Internet, such as Project Guttenberg ,the Internet Archive , and the Internet Children's Digital Library .

The Authenticity Puzzle

Since one of the teacher's major functions in the reading process is to motivate students and help them find good material to read, the availability of authentic texts at the appropriate level has always been a major piece of the reading puzzle. From very early on, the Internet has been an excellent source of authentic content. I see Internet-based materials at present as of three kinds:

(1) Text repositories, including digital libraries such as Project Guttenberg , are essentially storehouses for materials that would otherwise appear in dead-tree-and-ink format. Included in this category are most online journals, which essentially replicate the format of paper journals, with the addition of hyperlinked references and search capabilities. These types of texts are generally lengthy and appropriate for extensive reading. They are often best readby being downloaded and printed. Some sources for online libraries are mentioned above .

(2) Electronically mediated texts are "accreted" by authors, both amateur and professional, for their own purposes, whether informational, emotive, or propagandistic. These documents are "native" to electronic media, and include content Webpages, software, and electronic books. Theyare characterized by hypermedia and linking. They are intended to be read largely on screen and/or online. Web content pages generally target native speakers, although ESL/EFL and foreign language sites whose intention is language practice have been created by both teachers and students and are rapidly growing in number. Content Webpages appropriate for reading practice are discussed further in this section.

(3) Computer-mediated communications (CMC) offer interactive authentic language on bulletin boards, electronic lists, e-mail, chat, Weblogs (blogs), etc. Also native to electronic media, these genres diffe rin crucial ways from their print counterparts and present their own perplexities for the reader. CMC will be discussed in the next section.

Both government and non-profit organizations and commercial sites offer a wide range of educational Web-based activities makinguse of authentic (or semi-authentic) reading matter. For example, the Learning Resources site offers Web-delivered instruction for literacy level adults using CNN San Francisco news bureau stories. At other sites, lesson plans for reading online range from simple one-page text downloads to breathtakingly beautiful photographic imagesand tons of curricular support both in paper and online, as for example NASA Quest 's educational pages, mounted by the U.S.National Aeronautics and Space Administration, which help students participatein the exploration of space. The Discovery Channel, though a commercial site, has highly interactive, media-rich activities, for example, Celebrity Shark Week (see Figure 3 ), where the user is invited to take on the persona of a tiger shark.

Figure 3. Online activity from Celebrity Shark Week at the Discovery Channel online http://dsc.discovery.com/convergence/sharkweek2002/sharkweek2002.html.

Dozens of Websites publish content-based reading lessons that include what might be called "cognition-priming activities," such as the use of graphical organizers, Into-Through-Beyond activities, schema-building frameworks, vocabulary activation through media, online dictionary lookup on the fly (for example, VoCabulary links words on a Webpage to any one of several online dictionaries to provide point-and-click lookup; Babel Fish will translate a Webpage into a handful of different languages.), pre-reading questions, post-reading composition topics, and so on. The best of these activities generally are used to support the research-and-reading-based inquiry learning referred to generically as "WebQuest."

Bernie Dodge at UC San Diego defines a WebQuestas " an inquiry-oriented activity in which some or all of the information that learners interact with comes from resources on the Internet" ( Dodge, 1997 , p. 1). WebQuesting fits well with a constructivist approach to learning, with task-based and content-based activities, and with communicative learning theories, for it is usually performed as group-or team-based projects. The WebQuest Page ( Dodge with March, 2002 ) offers hundreds of examples created by teachers and organizedby U.S. school grade levels, adult literacy, ESOL, and foreign languages.Each project follows a carefully worked out, readily adaptable framework:introduce the problem, set a task and process, define resources, have students produce an end product, make explicit the evaluation criteria (see Figure 4 , left frame). Usually, the lower the level of student, the more the WebQuest materials are self-contained within the teacher's own Webpages. As an example, see An Insect's Perspective ( Tyson, 2000 ), a WebQuest for second graders, ages 10-11 ( Figure 4 ). In contrast, the more advanced levels are quite open-ended: students maybe asked to formulate their own research questions, and then are sent off to explore the Web (perhaps with a few suggestions as to appropriate search engines and sites to hit); learners develop their own responses to the research through group projects, suchas writing a paper, or creating a presentation or Webpage themselves.

Figure 4. Process page from an elementary school, An Insect's Perspective. ( Tyson, 2000 , http://projects.edtech.sandi.net/grant/insects/ .) Note the sequence of activities in the frame at left.

 Imagination Voyages ( Mills with Melin, 1998 ) is a good example of an ESOL WebQuest for advanced learners at an intensive English program in the U.S. WebQuests can involve a mix of print text (library research), physical research (for example,in the chemistry lab) and Web-based content, and they can be used to motivate students to learn more about a particular content area; science and social science WebQuests are especially numerous. Dodge holds regular text chats at the online community, TappedIn , where teachers can learn more about how to construct WebQuests appropriate to their own students. (As of this writing, Dodge is in the process of reorganizingthe WebQuest Page; new contributions are always welcome.)

 I refer to the abundance of authentic reading matter on the Web as the "authenticity puzzle" because authentic materials, those written for native speakers in the target language, are the goal for learners, but not always the medium through which that goal can be attained. In addition, materials written for young NS learners, for example fairy tales or comic books, may be at an appropriate reading level for beginners, but may be puzzling to those outside of the target culture, or offensively juvenile to adults. Because billions of documents now exist on the Web, ESOL teachers have been using technology to compensate for the difficulties of reading online, for example in the use of online dictionaries and translators, as indicated above. An interesting Internet-based corpora project, TextLadder(see Ghadirian, 2002 ), is underway in which textsare selected and assigned levels on the basis of the word frequencies of target vocabulary. Teachers or students may then select from targeted levels for various reading activities. The next section of this paper looks atsome other ways in which technology has offered both solutions to and new problems for the authenticity puzzle.

The Electronic Puzzle

One way to look at the electronic aspects of the reading puzzle isto compare the differences between reading print and reading electronically mediated texts (see Table 1). 

Table 1: Comparison of Print and Electronic Reading

Although print publications are clearly moving in the direction ofmore visual aids and more inventive layout, Webpages still have a tremendous advantage (or disadvantage?) in the amount of distracting material the reader will encounter. If we look at Celebrity Shark Week in full screen (not just the useable content, as in Figure 3 ), we see immediately what an inundation the senses undergo, even without the movement and color of the online original (see Figure 5 ).

Figure 5. Celebrity Shark Week with full complement of hyperlinked distracters, including EMicrosoft® Internetxplorer browser interface ( http://dsc.discovery.com/convergence/sharkweek2002/sharkweek2002.html ).

Not visible in this static illustration ( Figure 5 ) is the movie running in the dark square in the "introduction" section. The "Guilt Free Fast Food ad" is dynamic and flashes to new information. "Click here!" and "Go!" messages in bright colors beckon. The Thanksgiving message in the frame at right will change on each entry to the page. Thus, while the visual image is rich and alluring to the reader, i.e., motivating, it also presents numerous visual and cognitive distracters. One of the skills that advanced readers of Webpages quicklylearn (or try to learn?) is to ignore the flash and focus on content. Learners have more difficulty in shutting out, or even sometimes determining whatis an ad. Even advanced learners sometimes can become confused about whatis a legitimate learning path, as any user of the Internet will attest.In a recent study I conducted with experienced Internet users (Hanson-Smith,2003), 85% of respondents ( n=27) said they were irritated or distractedby commercials on the Internet. (Interestingly, only 2% of the respondentshad thought about whether ads were similarly distracting to their students, Hanson-Smith, 2003 .)

Hyperlinks are a blessing and a bane. They can jump us instantly from a text to a footnote or reference, to an online dictionary, to a picture or a movie, to another language, another country, or evenouter space--the possibilities are almost limitless and incredibly enticing. Roger Chartier in a recent online conference sponsored by the Bibliothèque Centre Pompidou declares the "electronic representation of writing radically modifies the notion of context, and as a result, the very process of theconstruction of meaning" ( Chartier, 2001-2002 ,p. 7). Put more forcefully, no text on the Web is the same for any two individuals, because the choice of following a link or not is based onthe reader's free will. While deconstruction as a literary critical movementhas tried to show us how the author's intentions may be hidden and perhaps ultimately unrecoverable, the task is additionally complicated in hypermedia. While the hypertext author constantly makes choices about presenting resources or structures of thought that would not fit into the accepted notion ofa printed text, the reader may choose or not to follow a link, to reconstruct the author's thought processes or not. Further, unlike traditional footnotes or references in printed texts, links function in many different ways. Harrison (2002) , describes seven different functions:authorizing (mailto, or About Us), commenting (opinions about a site, e.g.,press releases), enhancing (Site Map), exemplifying (specific examples ina broader category), mode-changing (Shopping Cart), referencing (bibliographyor appendix), self-selecting (For Seniors Only) ( Harrison,2002 , p. 7). The reader, particularly the learner, may not be fully cognizantof these varying purposes.

Managing chaos no doubt has tremendous value instretching us cognitively. However, hyperlinking can often extend us tothe limits. As teachers who write Webpages ourselves, we often try to offer aids to help students manage chaos; however we may end up making the reading acquisition process (a) slower and more inefficient or (b)far more complex than it need be. Much interesting research is now accumulating the role of hyperlinks in the reading process. To cite just a few: Al-Seghayer (2001) compared the effectiveness of video vs. still photos in teaching new vocabulary, finding in favor of the former. De Ridder's work (2002) seems to indicate that the increased clicking evident when visible (as opposed to hidden) hyperlinks were put into a text did not slow the reading process,but also did not increase the vocabulary learned: students liked to click,but the effects of instant lookup were not lasting. Perhaps because hyperlinkingcan be so quick (depending on bandwidth), there is less interference withthe comprehension process than with paper dictionary lookup. Altun (2000) , in a small pilot case study with two students' reading strategies, found that "navigating through links is acomplex cognitive activity in which various strategies are involved" (p.24), including a number of ways to circumvent poor design. An astute comment by one of the subjects was that "Getting lost is not the fault of usersbut the fault of designer" [sic]( Altun, 2000 , p.20). Both students wanted to write, that is take notes, as they read, and felt some frustration that current hypertext environments do not easily provide for this natural desire to create marginalia ( Altun, 2000 ,p. 24). Unfortunately, electronic books do not appear to solve this problem either.

Much still remains to be done in exploring the effects of hypermedia on learning. However, numerous articles describe the useof the Internet, multimedia, hypermedia, and computer-enriched activitiesas an integral part of good teaching practices. Bertelsen and Fischer (2002/2003) , for example, used multimedia scaffolding to support 10- and 11-year-old students' experiences with expository text in a social studies content-based setting. Yerrick and Ross (2001) report on the use of homemade computer-based video to help develop literacy in inquiry-based science instruction. A recent special issue of TESOL Journal on the theme "Constructing Meaning with Computers,"is a rich storehouse of ideas and plans based on constructivist principles.(See Healey and Klinghammer 's introduction to this issue for a handy definition of "cognitive" vs. "social" constructivism,2002, p. 3.) Researchers are in universal agreement that technology can motivate students, enhance authentic task-based education, and lead to autonomous learning. Nonetheless, how technology gets used--and how much--depends entirely on teachers' willingness to use a tool that directly diminishes their control over the instructional situation.

Another important element in the electronic puzzle is computer-mediated communication (CMC), which may include asynchronous bulletin boards or online discussion groups, student-student or student-mentor e-mailing, and synchronous chat. Scaffolding is an important element in both learning to speak and learning to write and hence, we may assume, in reading as well. E-mail has for the last decade been a favorite means to inspire students to write more, to write about topics that interest them, and hence to read what others have to say on those topics. As students read each other's work and communicate with more advanced learners or the teacher, they tend to use expressions others have formulated, thereby building the appropriatevocabulary to talk about specific subjects. (See Holliday,1999 , for a summary of the extensive research on student e-mail and language acquisition through output; and Peyton, 2000 , for an example of intranet chat scaffolding with Deaf students). More recently, Weasenforth, Biesenbach-Lucas, & Meloni (2002) have explained how threaded discussions on bulletin boards enhanced learning while producing some unintended consequences, as students on their own initiative took over and managed their own group discussions (p. 7). While these researchers felt that the asynchronous nature of bulletin board discussion was helpful, giving learners more time to read and formulate responses, other teacher-researchers have found synchronous chat to be even more motivating.

Real time chat has been called "talk through fingertips" ( Almeida d'Éça, 2002 , p. 5), and as it comes into wider use, students learn how to read and sort through multiple messages, interruptions of thought, and commingling threads of discussion--a very different kind of structure from either linear/paper or hyperlinked text, or threaded (categorized) messages on an electronic bulletin board--but one that is still manageable because a written record is maintained, and participants can learn ways to indicate the threads of conversation. Froman academic standpoint, the most useful chat is probably one focused on a series of discussion questions closely linked to recent classroom or online activities. Chat is especially effective in an environment like TappedIn , where breakout groups can isolate themselves as needed (thus cutting down on the confusion of multiple threads), and whereone can "project" a Webpage in a secondary browser window for all to lookat together. Nonetheless, this type of reading, though powerfully motivating,is often quite short and disconnected, sometimes overwhelmingly colloquial,and more like transcribed talking than other genres of written text. Almeida d'Éça (2002) provides feedback from EFL students on their first experience of the difficulties and rewards of chat:

here is really a little bit crowded, and the screen changing too fast, the conversation is hardly be focused on one certain topic. That's my opinion. (Grace1L);

and reading speed (WeiH);

you gotta think fast and type faster in Eng...good way to exercisemy brain (IrisCh);

it might be the first time and last time that I came here (CherryH);

I enjoy this new semester's "trend". . . cuz I love using computers...and it's definitely a must in the years to come (JoycePC).

(TappedIn log, 22 Sep. 2002, quoted in Almeida d'Éça, 2002 , p. 7.)

Text chat seems like a good introduction to real talk (unlike oral conversation, one can stop, scroll back, and read the previous comments), and it is yet one more piece of the electronic puzzle. Note WeiH's comment that "reading speed" is a consideration. CherryH and JoycePC express the two poles of chat: hate it or love it. Yet most users come to the latter point of view with just a little practice. Jewel Reuter (2002), a biology teacher, provides an excellent informal description of using  chat foracademic purposes, but the social elements of chatare perhaps its most impressive feature, particularly as the use of audio and video chat bbecome morecommon and the technology increasingly accessible.One teacher uusing textand voice chat simultaneously with students wrote with great feeling to the Webheads community about both the technical advantages of combined audio and text chat and the social benefits of online communities. Her remarks also include the sense of satisfaction the teacher feels in giving up control of the event (the quote reflects the informal, unedited nature of community discussion):

Last night I had a particularly rewarding session with a group of ESL learners, many of who are almost nightly attenders and who have therefore developed a lovely sense of shared community. The session revolved around different country maps which volunteers used to give mini-presentations using voice chat on places and things of interest in their countries. They also answered other participant's questions.Everyone was very encouraging of each other's efforts and the interaction between participants felt very authentic. Although it can be limiting tha tonly one person can speak at a time [using the audio channel], the text chat really makes up for this by allowing those who are not currently talking to give a running commentary on what is  being discussed. Those who lacked microphones were all actively encouraged by others to get them so 'we can hear your voice".   I really felt a sense of mutual respect and an openness to learn from each other.  I found it very easy to take a back seat role and let the students do most of the 'talking'.My role was just to keep things moving, make the occasional correction or clarification, to make sure that everyone was sent the web page under discussionand knew what was being discussed. . . .

Another feature . . . I really enjoy is being able to have the simultaneous voice and text chat. This makes it really easy to welcome and 'brief' newcomers without disrupting the flow of the voice chat discussion. It also allowsless confident learners to take a back seat and just contribute through the text chat or to have quiet 'asides' with others in the chatroom. Itsalso great from the teaching point of view because it allows you to make smooth corrections and modelling alternative ways of expressing ideas.I think what made last night really special for me last night was how different students each took the 'expert's' role when sharing their knowledge of their homelands. Although none of this stuff is probably new to all you'old hands' I'm still feeling that sense of wonderment about it all taking place in real time even though we are all spread across all corners of the globe. At one point, a young Polish lady was talking about her 'place' and her dog started barking loudly in the background. This created a lot of amused commentary from participants because it was such a homely sound and seemed like it was coming from just around the corner! It really underlined for me the feeling of participating in a 'borderless' community. "Sonja," evonline20002_webheads message posted December 12, 2002;also cited in Stevens, 2002 .

Solutions to the Puzzle

In a cardboard jigsaw puzzle, the solution to how pieces fit sometimes comes in a sudden, brilliant, unpremeditated flash. The solution to the electronic reading puzzle is far more difficult andtime-consuming. This section offers some suggestions as to how we can better address our learners' need to read electronic documents.

One piece of the solution to our puzzle must be to create standards for Websites that will be accessed by language learners.The design of Webpages must take into account accessibility for specialneeds students, organizational patterns of the target language, the appropriateuse of visual and auditory linking, and the building of cognitive and metacognitive skills. Among many organizations working on standards for technology ineducation are CAST (Center for Applied Special Technology); CETIS (The Centre for Educational Technology Interoperability Standards; see Learning Technology Standards, 2002 ); ISTE (International Society for Technology in Education; see ISTE NETS Project, 2000-2002 ); IATEFL (the International Association of Teachers of Foreign Languages, a mainly European-based group); TESOL (Teachers of English to Speakers of Other Languages); and other teacher-based organizations. (To participate in the TESOL technology standards effort,contact the author.) Since users evaluate Website credibility largely onthe basis of visual content, rather than on the breadth, depth, and qualityof information (see Fogg, et al., 2002 , p. 2), critical thinking must play an important role in learner and teacher standards.

The importance of critical or "inquiry" thinking (also referred to as constructivist education, holistic learning, et al.) cannot be underestimated as part of the solution to the electronic reading puzzle. Educators in the U.S. have long noted the "fourth-grade slump" where readers seem unable to parse meanings out of expository text (see Beck & McKeown, 1991 ). Until the fourth grade, reading instruction often mainly focuses on decoding skills and assumes comprehension will take care of itself. Some students never make the transition. Themeans to move to inquiry learning are well exemplified: Brandl, (2002) offers a paradigm leading from teacher-directed tasks to student-directed constructivist learning. Crandall et al. (2002) summarize the use of cognitive strategies for literacy, while Florez (2000) assures us that even literacy-level learners use strategies such as predicting, self-monitoring, and, interestingly, social interactions and "coping" strategies: risk-taking, willingness to laugh at one's own mistakes, validating and reassuring others, and so on( Florez, 2000 , pp. 2-3), the very strategies mentioned in Sonja's description of chat quoted extensively above .

Teacher education itself will have to place even greater emphasis on the importance of inquiry as opposed to simple fact-finding. Green & O'Brien (2002) report that generally teachers set students to tasks directed at "finding answers to the questions rather than evaluating the quality of the information" both in off-line and online assignments (p. 1). They recommend constructivist practices that will help teachers produce "active students facing cognitive challenges" (p. 51). This approach must also be used in educating reading teachers.Two excellent articles, Boxie & Maring (2001) ,and Maring, Levy & Schmid (2002) , describe "cybermentoring" as a means of delivering literacy tutorials to K-12 students while training new teachers; activities they suggest include e-mail, video conferencing, and instructional Webpages. Perhaps the most interesting notion of ongoing teacher education can be found in the "communities of practice," as developed in the Webheads in Action group, described in Stevens (2002) , Stevens (2002-2003) , and Steele (2002) . For a general discussion of communities of practice, also see Wenger (1998) .

One means to bridge the gap between decoding and comprehension or skill-building vs. inquiry learning, may be to move gradually from heavily teacher-supported Websites that model and guide inquiry thinking, to student-centered authentic, content-based tasks, what Brandl calls "learner-determinedlessons" ( Brandl, 2002, p. 93). Certainly, teachers must be made aware of the differences between teacher-controlled vs. student-controlled approaches. However, another paradigms to allow students (and teachers) from the very beginning to explore technology and find their own meanings by using it. An extreme example of this model is the "Hole in the Wall"experiment in "minimally invasive education" taking place in India (see Mitra & Rana, 2003 ; and a recent report on PBS's FRONTLINE/World, 2002, with further documentation at Kid's-Eye View, 2002 ). Two educator-researchers at the Cognitive Engineering Research Centre constructed a well protected outdoor kiosk containing an Internet-linked computer next to a New Delhislum. Children were allowed to explore the computer hands-on, teaching themselves how to use its drawing software and browsing the Internet. Children in the 6-12 age group "fiddled around" with the touchpad, and accidentally found that it was clickable; they invented their own terms for the objects and events they saw, e.g., "needle" for the cursor, and "Shiva's drum" for thehourglass "busy" symbol. Within a fortnight, the children had learned (and taught others) to make short-cuts, create folders, use the calculator, and eventually maximize and minimize windows, read the Hindi Times online, andeven change the wallpaper setting. The first of the kiosks was placed in the wall of the Research Centre's compound, hence the name "Hole in the Wall,"but over 50 others have now been erected in a variety of settings where disadvantaged kids may construct their own learning. The holes in the wallgive new force to the idea that children learn best through experiment, trial-and-error,human interaction, and the exploration of meanings significant to themselves--in other words, the constructivist approach. Teacher education in technology can proceed as well along similar lines. (See Hanson-Smith, 2000 , where teachers in small groups are turned loose on computers to create presentation programs with only minimal instruction).

We may well contemplate what is at stake in teaching students good electronic reading habits. Steve Lawrence of the NEC Research Institute offers us some sobering statistics in a recent article in Nature ( Lawrence, 2002 ): using the Research Index, he cross-referenced 119,924 refereed articles read at conferences in computer and related scientific disciplines and found a "clear correlation between the number of times an article is cited, and the probability that the article is online" (p.1). In fact, there was an average of 336% more citations to online articles than to offline articles across 1,494 publication venues (p.2). Lawrence goes on to speculate that the free online availability of articles, online archives, direct connections between scientists and research groups, hassle-freelinks, indexing search engines, and so on, can "maximize impact, minimize redundancy, and speed scientific progress" ( Lawrence,2002 , p. 3). Indeed, even in the field of ESOL and literacy education,articles that are offline are beginning to be increasingly invisible. Students around the world, whatever their discipline, must be able to use electronic media efficiently and resourcefully, and the electronic reading puzzle must be one of the areas where teachers become skilled mentors and guides to solutions.