Do computers make us smarter? Probably not. But they can reduce the burden of some largely mechanical processes through which we develop, refine, and express ideas—which is a lot of what it means to think. The clearest example is word processing, despite the ugliness of that term. No one believes that you become a better writer just by sitting at a computer. But it would be even more far-fetched to imagine that the olden-days burdens computers now spare us—for instance, retyping pages or whole reports to accommodate a minor change—were in any way a plus.

In a surprisingly wide range of other ways, the simple, brainless efficiencies offered by computers can assist in the tasks that make up intellectual work. Before considering one especially ambitious new offering of this sort, it’s worth reviewing the practical, chore-like components of high-level modern work and the corresponding programs that, in my view, handle each chore best.

An elementary step is capturing thoughts—ideas, obligations, possibilities—when and where they occur to you. This is so you don’t forget them and, more important, so you’re not always wondering what you might have forgotten. For life away from the computer, the small notebook in the pocket, or the stack of index cards, is the ideal note-taking tool—as long as you have ingrained the habit of taking these out of your pocket and doing something with them when you get home.

For when you are at the computer, a good note-taking program should be “pop-up” or “modeless,” which means that you can immediately call it up, usually with one keystroke, without switching away from the Web page you are viewing or the document you are working on. The program should also make it easy for you to sort through and deal with those quick notes when you review them later on. The pop-up feature first got wide notice in the early 1980s, thanks to Philippe Kahn’s pioneering Sidekick program. I have used many similar PC utilities since then, but have relied for most of the last decade on the pop-up “Zooter” feature in Zoot, a data manager from Zoot Software. GyroQ from Gyronix, mentioned here recently, has similar features, as does VoodooPad for the Mac.

The next largely mechanical task is saving material you come across in your work, whether it is something unexpected on the Internet or the result of more purposeful research. There are countless tools of this sort; the one I now use is OneNote 2007 from Microsoft. I like it because it can handle almost any kind of information—Web clippings, PDFs, audio or video files, straight text—and index it for quick retrieval. It also has an elegant feature that makes capturing information utterly painless. When something you want to save is on your computer’s screen, you can press a button or two and “print” that blog posting—or photo, or e-mail, or online receipt—to your OneNote file. It’s like storing paper documents in folders, except that it’s faster, easier, and more reliable when you look for the material later on. Microsoft has made other Office programs available for the Mac, but not yet OneNote. Scrivener, a new research and writing application, is what I would try on the Mac.

The next practical task involved in thinking is finding things when you want them—the right citation for your legal argument, the right chronology to remind you who said what when. Mac users have an advantage here: In addition to the Spotlight indexer built into OS X, DevonThink (mentioned here before) has its own powerful “fuzzy search” tools to help you find things even if you’ve misremembered the name or keyword you’re looking for. For PCs, the free search utility X1, from X1.com, is unsurpassed.

Next is sorting, the important and subtle task of grouping items according to similarities and differences. In a sense, this kind of pattern recognition is the highest level of human intelligence. Seeing patterns where others see confusion is what distinguishes the great chess masters—and the great diagnosticians, historians, philosophers, legal theorists. Recognizing distinctions is the other side of the same skill.

Such compare-and-contrast judgments are extremely difficult to produce through computerized intelligence. But computers can help people exercise their own judgment by providing cues, props, and reminders. They do this mainly by allowing you to apply “tags”—identifiers that help in later assessments of data. For 10 years I have used Zoot, even though it is idiosyncratic and somewhat hard to learn, because it is more powerful and supple in such tagging than anything else I’ve seen. If I read an article, I can decide that a certain paragraph is relevant to China—and to a meeting I’ll have next week, and to something I want to see in India. With Zoot I can assign it to all those categories with a few keystrokes. The computer’s role here is to provide a record of my judgments about similarities and differences, so I don’t have to try to keep them in my head.

Both sorting and the next step, outlining, take us closer to the point where mechanical processes merge with intellectual ones. Assigning something to a category inevitably affects our conception of that category, and arraying ideas visually, as in an outline, inevitably affects our view of how the ideas fit. The most straightforward computer outliners are cleaner, faster ways of creating hierarchies than is normally possible using pen and paper. NoteMap, now owned by the LexisNexis company, is an elegant PC version of the outlining techniques we learned in high school; OmniOutliner is its counterpart for the Mac. Most of the time I end up using BrainStorm for outlines on the PC: it is quick and flexible and has features the others lack.

But after a while the plain old I, A, 1, a–style outlines seem pointless on a computer. They are one way of depicting the connections among facts or concepts, and were the most feasible way when we were working with typewriters. There are many other ways to represent these relationships visually, and many fascinating programs help users explore these possibilities. Personal­Brain, for PCs, shows ideas or items as connected points in a sort of 3‑D representation of space. You can see how it works at TheBrain.com. MindManager, for PCs and Macs, and its free counterparts like FreeMind apply the popular concept of “mind mapping” to show how ideas mesh and diverge. The Axon Idea Processor, the product of more than 15 years’ solo labor by an engineer named Chan Bok in Singapore, offers an unbelievable array of PC tools for visualizing information: flow charts, fish-bone diagrams, and several dozen more. It’s at tinyurl.com/y3gg88.

This leads to the newest ambitious entry: Rationale, an “argument processor” from a start-up company in Melbourne, Australia, called Austhink. The firm’s CEO, Tim van Gelder, is a former academic philosopher whose specialty was teaching critical thinking—that is, preparing students to examine the premises of any argument, another person’s or their own. He had a discouraging experience in the 1990s when teaching such classes at universities in the United States. “Despite my best efforts, and maybe theirs, it just wasn’t working,” he told me in a Skype conversation, he in Melbourne and I in Shanghai. He was gracious enough not to attribute this failure to the defects of America’s K-12 school system. Instead, he concluded that people in general needed better training in assessing arguments. After returning to Australia, he raised money to start a company and create a program that could be used by schools for teaching logic.

In operation, the Rationale program is quite simple. You state a main contention you are trying to test—I should buy a new house, we should invade Iran—and then systematically list each of the supporting claims for it. Then you list the objections to each claim, and the rebuttals to those objections, and so on until you’re down to first principles—all of which are shown as connected boxes on a map. “To the extent you are perfectly clear about your thoughts, this should be a trivial process,” van Gelder told me. But in reality, he said, people find it more challenging than they expect, and this visual representation of the claims and counterclaims generally provokes a new perspective on the ideas at stake.

The more factors there are to weigh in making a decision—and, especially, the more views there are to reconcile when more than one person is involved in a choice—the more helpful this logic map can be. For example, a “tree” view in Rationale can show the full chain of assumptions that lead to a particular conclusion, which in turn helps identify exactly where people with different views disagree. “Everyone knows that complex structure is generally more easily understood and conveyed in visual or diagrammatic form,” van Gelder wrote in an academic paper. “That is why, for example, we have street maps rather than verbal descriptions of the layout of cities.” The same principle applies in complex debates, he told me, because in all but the simplest discussions people have a hard time remembering all the relevant considerations.

Van Gelder’s initial sales target was schools and universities, but he increasingly sells to consulting firms, govern‑ ment agencies, and other groups wrestling with decisions, as well as to individuals. The strongest interest has come from U.S. intelligence agencies, which are using the software to train analysts to think critically about intelligence claims. I was gracious enough not to ask van Gelder why he didn’t finish the program a few years earlier.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.