Photo: Intel Free Press / Wikimedia Commons / CC BY 2.0
Use library databases to find books and articles.
That answer sounds plain, but the important part is the split. A library catalog is built to locate books, journals, and other items a library owns. A database is often built to find articles, chapters, reports, and other indexed content. The two overlap sometimes, but they do not do the same job.
I keep this distinction close because many search problems begin with a bad starting place. If the question is about a book, a catalog usually matters first. If the question is about current scholarship, a subject database usually does more useful work. The fastest search is not the one with the most results. It is the one that matches the type of source.
The phrase “library databases” covers more than one thing. Some databases point to citations and abstracts only. Others give full text. Some are broad and catch many fields. Others are narrow and focus on one subject, like history, literature, or education. That range matters, because coverage shapes what can be found. A result list can look rich and still miss a whole body of work if the database does not index it.
Books and articles also behave differently in search. Books tend to gather ideas and frame a field. Articles tend to report narrower findings and arrive sooner. For that reason, a search for recent work often leans on articles. A search for background, theory, or long-term scholarly debate often leans on books. I do not treat either one as superior. I treat them as different kinds of evidence.
The search box itself can mislead. A simple keyword search may pull useful items, but it may also pull too much noise. Many databases allow subject terms, author names, title words, date limits, and source type limits. Subject terms are controlled words assigned by indexers. They help when a topic has many names or spellings. They are not always visible at first glance, but they can narrow a search in a real way.
This is where library databases become more than a pile of files. They are structured tools. A good database tells me what it covers, how far back it goes, and what kind of record it gives for each item. If that is not clear, I am cautious. I do not assume that a search result means broad coverage. I assume only that the item is in that database’s index, or in its full text, if the record says so.
The catalog deserves special care. It is often the best place to find print books and ebooks held by a library. It may also list journals, but not the articles inside them. That limit is easy to miss. A journal title can appear in the catalog, while the article a researcher really wants sits only in a subject database. That is one reason a search can feel empty when the wrong tool is used.
The reverse also happens. A database built for articles may include ebooks or some book chapters, yet still miss many print books. So the best question is not “Which one is better?” It is “What kind of source am I trying to locate?” That question keeps the search honest and saves time.
I also watch for uncertainty in the record itself. Database coverage can change. Interface labels can change. A platform may say it searches “books,” but that claim may mean ebooks, chapter records, or just selected collections. Publisher language is useful, but it is not the same as a full map of content. When the documentation is thin, I mark the limit and move carefully.
This is the part that matters most for a reader who wants books and articles, both. Use the catalog for books and owned items. Use library databases for articles and deeper subject searching. Then use both together when the topic calls for it. The goal is not to collect every possible result. The goal is to know which tool is doing which job.
That plain division is the best search habit I know. It keeps the work clean. It also keeps expectations realistic, which is often the same thing as being accurate.
The Source List fits that same standard: one digital source worth knowing, one search tip, and one honest limitation. That is the right shape for this topic, because the search is only useful when its limits are named plainly.