A typical paragraph is likely to rely on both generalizations and particulars. A generalization is a broad statement that establishes a point, viewpoint, or conclusion. A particular is an instance, a detail, or an example — specific evidence that a general statement is reasonable. Your particulars support your generalizations; compelling instances, details, and examples back up your broader point. Likewise, your generalizations pull together your particulars, identifying patterns or connections that relate individual cases.

To link particulars and generalizations, you can use an inductive or deductive process. An inductive process begins with the particulars — a convincing number of instances, examples, tests, or experiments. Together, these particulars substantiate a larger generalization. In this way a number of long-term studies of weight loss can lead to a consensus about the benefits of walking, eating vegetables, or some other variable. Less formal inductive reasoning is common as people infer or conclude that particulars do or do not support a generalization. If your sister ate strawberries three times and got a rash each time, she might infer that she is allergic to them. Induction breaks down when the particulars are too weak or too few to support a generalization: not enough weight-loss studies have comparable results or not enough clear instances occur when strawberries — and nothing else — trigger a reaction.

A deductive process begins with a generalization and applies it to another case. When your sister says no to a piece of strawberry pie, she does so because, based on her assumptions, she deduces that it, too, will trigger a rash. Deduction breaks down when the initial generalization is flawed or when a particular case doesn’t fit the generalization. Suppose that each time your sister ate strawberries she drizzled them with lemon juice, the real culprit. Or suppose that the various weight-loss studies defined low-fat food so differently that no one could determine how their findings might be related.

Once you have reached your conclusions — either by using particulars to support generalizations or by applying reliable generalizations to other particulars — you need to decide how to present your reasoning to readers. Do you want them to follow your process, perhaps examining many cases before reaching a conclusion about them? Or do you want them to learn your conclusion first and then review the evidence? Because academic audiences tend to expect conclusions first, many writers begin essays with thesis statements and paragraphs with topic sentences. On the other hand, if your readers are likely to reject an unexpected thesis initially, you may need to show them the evidence first and then lead them gently to your point.

In “The Good Heart” (Newsweek 3 Oct. 2005), Anne Underwood opens with a paragraph organized inductively: she describes a particular situation that helps substantiate a broad, even surprising, generalization.

You can call it the Northridge Effect, after the powerful earthquake that struck near Los Angeles at 4:30 on a January morning in 1994. Within an hour, and for the rest of the day, medics responding to people crushed or trapped inside buildings faced a second wave of deaths from heart attacks among people who had survived the tremor unscathed. In the months that followed, researchers at two universities examined coroners’ records from Los Angeles County and found an astonishing jump in cardiovascular deaths, from 15.6 on an average day to 51 on the day of the quake itself. Most of these people turned out to have a history of coronary disease or risk factors such as high blood pressure. But those who died were not involved in rescue efforts or trying to dig themselves out of the rubble. Why did they die? In the understated language of the New England Journal of Medicine, “emotional stress may precipitate cardiac events in people who are predisposed to such events.” To put it simply, they were scared to death.

Underwood reviews the impact on heart attack patients of various factors such as anxiety, depression, and childhood trauma. Then, in the next passage, she states and supports a generalization about effects of common stresses in adult life, citing the results of an inductive study. In the second paragraph, she deductively applies the generalization to a particular case.

And if stress in childhood can lead to heart disease, what about current stressors — longer work hours, threats of layoffs, collapsing pension funds? A study last year in the Lancet examined more than 11,000 heart-attack sufferers from 52 countries and found that in the year before their heart attacks, patients had been under significantly more strains — from work, family, financial troubles, depression, and other causes — than some 13,000 healthy control subjects. “Each of these factors individually was associated with increased risk,” says Dr. Salim Yusuf, professor of medicine at Canada’s McMaster University and senior investigator on the study. “Together, they accounted for 30 percent of overall heart-attack risk.” But people respond differently to high-pressure work situations. The key to whether it produces a coronary seems to be whether you have a sense of control over life, or live at the mercy of circumstances and superiors.

That was the experience of John O’Connell, a Rockford, Illinois, laboratory manager who suffered his first heart attack in 1996, at the age of 56. In the two years before, his mother and two of his children had suffered serious illnesses, and his job had been changed in a reorganization. “My life seemed completely out of control,” he says. “I had no idea where I would end up.” He ended up on a gurney with a clot blocking his left anterior descending artery — the classic “widowmaker.” Two months later he had triple bypass surgery. A second heart attack when he was 58 left his cardiologist shaking his head. There’s nothing more we can do for you, doctors told him.