CALICO SOFTWARE PRODUCT REVIEW
CALICO Journal, Volume 14 Number 2-4, pp. 165-171
1. General Description
Product at a glance
Product type: Concordancer
To date the design of concordance programs has been characterized by limitations. Concordance programs that work well enough for researchers attuned to working within the limitations of DOS-based interfaces can prove frustrating, even daunting, for casual student users (not to mention their teachers). Two limitations have been in processing power and fluidity of interface. Processing power has of course rampantly increased in small, affordable computers available to education, and on the PC side, Windows operating systems have enhanced the intuitive use of these computers. Monoconc is the first commercially available concordancer for PC take advantage of the power and fluidity of the newer machines and interfaces, and as such it is a tool which brings the potential of concordancing to language learning that much closer to being widely realized.
The accompanying manual for Monoconc is excellent. It begins with the briefest of overviews of how to run the program, and goes on to reveal more of the program features in complete, clearly written detail. According to the author, the documentation addresses two audiences, those who would like to know how to use concordancers with students, and those who would like to know how to use this particular one.
For people in the former group, the emphasis here is on considerations in assembling corpora. There are not many suggestions about how to use concordancers with students beyond a few examples, but users desiring such suggestions can readily find them elsewhere (e.g. Johns and King, 1991; Stevens, 1995; Tribble, 1989 and 1990; Tribble and Jones, 1990).
For people wishing to use this concordancer, the manual isn't really necessary (unless you aren't all that familiar with concordancers or with Windows, in which case you will be pleased to find that the explanations don't assume much prior knowledge). There are only half a dozen top-level commands which pull down to 2 to 6 sub-level commands each, which may in turn generate dialog boxes, but these are again intuitive if you know what sort of information you are looking for from a concordance program. Without glancing over the manual, it would be possible to not be aware of the many complex features available in what is on the surface a simple program. For example, I found the manual useful in helping me to discover how I could build a corpus from files scattered all over my disk and augment or delete files from my corpus at will.
The manual includes a complete listing of all commands both in alphabetical order and as an outline arranged in the order of the pull down menus. The alphabetical list is annotated with a brief explanation of what each command does. The index to the manual seems adequately utilitarian, and its creation must have been greatly facilitated by use of the program itself. Thus information is easy to find in the manual.
Monoconc provides the usual concordance features: a KWIK display on a word or string which can contain 3 kinds of wild characters (* = any number of characters, ? = 1 character, and % = 1 or no characters), and output can be sorted on one or two words left or right. In addition, Monoconc allows users control over a number of parameters; for example, limits on the number of hits, adjustment of the amount of context of which the target is the center, treatment of case, and what characters besides spaces to treat as word boundaries.
As I worked with the program, I found many of the features on my personal wish list to be included. Users are granted the ability to recall and alter previous search patterns. There is the ability to easily halt searches in progress and get results thus far (a feature which, when lacking, can lead to frustration in student users). The program will work on a variety of language-specific fonts (the manual claims a degree of success with Arabic and Thai). And with "Search Append" there is an ability to build a concordance of all forms of a lemma (e.g. go, went, gone, going) by appending output to previous searches and deleting from the list in the case of false hits (e.g. deleting a hit on "gong" from an output of go* to get "go", "going" etc.). The only thing missing from this concordancer is a corpus.
One FAQ on the various lists geared to teachers of languages regards where programs can be obtained to get lists of words in a text and their frequencies. This capability is included in this concordancer, where collocation and word frequency data are at your fingertips. After your first concordance, you will know how many words are in your text, and you can have a list of those words arranged either by frequency or alphabetically with data on the number and percentage of times each appears in the text. You can summon a list of all words or just content ones, or you can get a list of all words not on your own stop list, which you can either create on the fly or load in from a text file (a "stop list" is a list of words such as 'the' or 'a' which can be ignored for the purpose of gathering word frequency data). With equal ease, you can have a list by frequency of collocates (up to two words left and right) of any string you concordance.
Students using concordancers are often frustrated by spelling errors and by an inability to predict what the lemma should be when they know one form of a word. Therefore, one suggestion I might make for improvement of the program would be to allow users to search and block off bits of text displayed in the "Corpus Text" window and then paste this to the search string dialog box. Another feature I would appreciate is knowing for each hit what file that instance of text can be found in. One advantage of this program is that it lets you select files by browsing devices attached to your computer and in so doing easily construct a concatentated corpus, but as the program stands, it does not let you know which of those files you should look in once you have detected a string, in case you would like to go there for more information. This information could be easily provided, I would think.
Speed Reliability Compatibility
The program is fast, reliable, and compatible with a range of low to high end lap and desk top computers running Win 3.1 to Windows 95. Your ability to access the full range of data available may be restricted by the capabilities of your computer, but limitations noted in beta versions of the program have largely been worked out. In early versions of the program, I was not able to get frequency data on words that occurred only once in large corpora, but in the latest version provided me, I was able (in win 3.1 on my 8 mb laptop) to get a frequency count on all words, even those occurring only once, with no stop words in effect, for a 179,215-word text. I used a 248,800,000 byte file (books94.m11) on my Microsoft Bookshelf cd-rom to have Monoconc run a concordance on the string "thespian" which I had previously determined to occur not too often in the file. It took all of 20 minutes on a pentium 166 with 32 meg of ram running Win 3.1 (17 min in Win 95), but Monoconc gamely muddled through and reached the end of the file without falling over in a heap. I was impressed.
Screen management, Navigation, User interface and Exploitation of computer potential
Monoconc has variety of very useful features, a lot of them having to do with the fact that this is a Windows-based product, and so, for example, you have the ability to find files by browsing and delimit them (e.g. to show only files with the extension "txt") and select a number of them by clicking individual ones or dragging through the list. You can integrate concordancing with other things you might be doing; for example, you could be examining a set of texts for inclusion in your curriculum and have the concordancer running in a separate window, helping you to analyze those texts and isolate those most suitable to your syllabus. You can keep multiple windows of concordance data active and size them or close them at your convenience. You can save concordance and collocation data output to files and manipulate them in a word processor external to the concordance program (to put word frequency data in a report, or create a concordance-based exercise as a handout for students).
One disadvantage of a Windows interface is that the default font that came set with the versions of the program I tested did not allow the target strings to line up well in a KWIC display. Setting the font to Courier solved the problem, and might be advisable for a default setting in case this doesn't occur to every user.
Concordancing falls toward the collaborative, facilitative end of the spectrum of activities that it engenders. Much has been written on concordancing as a tool for language learning (Higgins, 1991; Johns, 1988; Johns and King, 1991; Stevens, 1995; Tribble and Jones, 1990) but briefly, a concordancer is a tool to allow learners as well as researchers to investigate patterns in a language. The test of a good concordancer is not only the range of appropriate explorations possible, but how quickly and easily and to what extent the operations can be carried out.
In these tests, Monoconc compares exceedingly well with other concordancers. In addition to most of the operations one would expect such as keyword in context (KWIC) displays of target strings, sortable up to two words left and right, and generous control over how results are culled, organized and filed, Monoconc provides useful statistics on collocates plus number of words in the text and their frequencies. In performing its operations, the software is fast and its interface is intuitive. Activities possible with the wealth of resulting information are legion. Many such activities are mentioned in the references cited below, which in turn cite other references.
Teacher fit (Approach)
Regarding teacher fit, a case can easily be made to reconcile concordancing with modern teaching methodologies. A good example is found in Johns's (1991a and 1991b) characterization of the role of concordancing in data-driven learning. However, preference for concordaning both from a teacher's and learner's standpoint may in the end be a question of style. For a teacher, use of concordancing with students increases exposure to unpredictability of outcome, and student appreciation of concordancers is limited to the more inductively inclined, the more tolerant of ambiguity, and those best able to cut through a bramble of extraneous data to find the berries in the bush. Even where individual preferences may discourage use of concordancers, it can be argued that some training in their use inculcates strategies desirable in "good" language learners and teachers.
In summary, I found this to be an excellent program, and possibly the best of its genre. It is software capable of serving the most demanding of concordance enthusiasts as well as acting as a basic practical tool for any language learner or teacher.
As a mild criticism, I might mention that the sort options in this program, while adequate for most operations the user is likely to need, fall slightly behind those in other concordancers. I like, for example, Microconcord's ability to sort instantaneously three words left or right by pressing a Ctrl-arrow key, and the way by which that program calls attention to the word sorted on by highlighting it. With Longman Mini Concordancer, I like the proximity search, which allows you to locate collocates within a certain distance of your target string. In Monoconc, you can approach proximity and third, forth etc. word-over searches by concordancing on "targetstring * * thirdword," (with 2, 3, ... asterisks). This is slightly more cumbersome than with the other two programs, and for more than two-words-over searches, the chance of serendipitous discovery is lost.
However, in performing sorts, Monoconc has the most intuitive interface. LMC requires you to pull down a separate menu to do a proximity search (so that unsophisticated student users tend not to). With Microconcord, I tend to forget the magic ctrl-arrow combination, which is not indicated on the screen, so that I often have to look up how to sort in the on-line documentation before I can actually do it.
Ease of use (student/teacher): The program is very easy to use.
Value for money: A concordancer is a must in any institute where languages are studied by means of computers, and the price for this software seems reasonable considering its quality.
For more information see References below.
4. Technical details
Supplemental software requirements:
Price Individual copy: $89 Site license: $425
5. Reviewer information
Vance Stevens is CALL Coordinator for AMIDEAST's MLI Project in Abu Dhabi and has 20 years experience teaching most aspects of ESL in Texas, the Middle East, and Hawaii. Vance has published widely in the field and engaged in numerous research projects. Many of his projects have involved concordancing as a tool for language learning.
AMIDEAST - UAE/MLI Project (425 0300 13 0)
1730 M St. NW, Suite 1100
Washington DC 20036-4505
Finding out about Concordances on the Web
A good starting point for comprehensive information on concordancing is Catherine Ball's Tutorial: Concordances and Corpora at www.georgetown.edu/cball/corpora/tutorial.html
See also Michael Barlow's Corpus Linguistics page at www.ruf.rice.edu/~barlow/corpus.html
Another interesting source for information on concordancing is Tim Johns Data-driven Learning Page at web.bham.ac.uk/johnstf/timconc.htm
This review ©1997 CALICO, Computer Assisted Language Instruction Consortium