01 · The Question
Why Do Databases Offer Both Keywords and Subject Headings?
You are searching an academic database for research on a concept such as workplace stress, online learning, or childhood obesity. You enter the words that seem obvious, retrieve some useful articles, and then notice that the database also offers terms labeled Subject Headings, Subjects, Descriptors, Index Terms, or perhaps MeSH Terms.
That raises a practical question: if your keywords already describe the topic, why would you need another set of terms?
The distinction matters because keywords and subject headings approach retrieval differently. One searches the language appearing in or associated with records. The other uses standardized concepts assigned through a database's indexing system. For a quick exploratory search, that difference may not seem important. For a careful literature search, it can substantially affect what you find and what you miss.
03 · What You Need to Know
How Keyword Searching and Subject Heading Searching Actually Differ
What Is a Keyword?
In everyday searching, a keyword is a word or phrase that you enter because you expect it to represent the concept you want to find. If your topic concerns academic burnout among university students, you might search terms such as academic burnout, student burnout, college students, or university students.
This is often called free-text searching or text-word searching, although database terminology varies. Depending on the database and the field you select, your terms may be searched in titles, abstracts, author-supplied keywords, or other searchable fields. This is why it is useful to understand which database fields your search is actually examining.
The defining feature is that you are searching using language that appears in, or is associated with, the bibliographic record rather than relying exclusively on a standardized indexing term.
What Is a Subject Heading?
A subject heading is a standardized term used to represent a subject or concept in an indexed record. The National Library of Medicine defines subject headings as terms or expressions that provide a major means of subject access to bibliographic records. Subject headings usually belong to a controlled vocabulary: an organized set of authorized terms used consistently for indexing and retrieval.
A well-known example is Medical Subject Headings (MeSH), the controlled vocabulary produced by the U.S. National Library of Medicine for indexing, cataloging, and searching biomedical and health information. Other databases may use their own thesauri, descriptors, or subject-heading systems. CINAHL, for example, uses CINAHL Subject Headings, while APA PsycInfo uses terminology from the APA Thesaurus of Psychological Index Terms.
Keyword
A word or phrase selected by the searcher to represent a concept and searched in one or more database fields.
Subject heading
A standardized term from an indexing vocabulary used to represent the subject of a record.
Why Does the Difference Matter?
Researchers do not always use the same words for the same concept. One author may write heart attack, another myocardial infarction, and another may use a more specific clinical term. Terminology may also differ across disciplines, countries, periods, and research traditions.
Keyword searching responds to this variation by allowing you to search the different expressions themselves. This is why identifying the different terms researchers use for the same concept is an important part of building a search strategy.
Subject indexing takes another approach. An indexing system assigns a standardized concept to records judged to concern that subject. In MeSH, for example, the preferred descriptor Myocardial Infarction provides a consistent subject access point even when authors use different wording. NLM explains that MeSH is designed to provide consistent retrieval across documents that may use different terminology for the same concepts.
Keywords Depend More Heavily on the Language Used in the Literature
Suppose you search only for online learning. Relevant authors might instead use e-learning, web-based learning, virtual learning, online education, or another expression. A literal keyword strategy can miss records when the wording you chose does not occur in the fields being searched.
This does not make keywords unreliable. It means that keyword searching usually requires you to think in terms of synonyms, alternative expressions, acronyms, spelling variants, and changes in terminology. Those terms can then be connected appropriately within a broader academic database search strategy.
Subject Headings Can Gather Different Expressions Under One Concept
The strength of subject headings is conceptual consistency. If records using several different expressions have all been indexed under the same authorized heading, a search for that heading can retrieve them without requiring you to predict every expression used by every author.
This is one reason controlled vocabularies exist. NLM describes controlled vocabulary as a specified list of terms intended, among other purposes, to prevent related subjects from being scattered across different headings.
Subject headings may also be organized hierarchically. MeSH, for instance, arranges descriptors from broader to more specific concepts. PubMed normally includes more specific MeSH terms beneath a selected heading when that heading is searched, although searchers can turn this feature off. The exact behavior depends on the database and platform, so you should not assume that every subject-heading system operates like PubMed.
Subject Headings Are Database-Specific
A keyword such as teacher burnout can be tried in many databases. A subject heading cannot necessarily be transferred in the same way.
Different databases may use different controlled vocabularies, different preferred terms, different hierarchical relationships, and different indexing policies. Even when two databases cover overlapping literature, the appropriate subject heading may differ. CINAHL Subject Headings, for example, follow the structure of MeSH but are maintained for CINAHL's subject indexing and include terminology relevant to its own coverage.
This becomes particularly important when searching multiple databases. You can preserve the underlying concepts, but you may need to translate the search strategy for each database rather than copying the exact syntax and subject headings unchanged.
| Feature |
Keywords |
Subject Headings |
| Where terms come from |
You identify words and phrases used to express the concept. |
You select authorized terms from the database's indexing vocabulary. |
| What they primarily capture |
Language appearing in specified searchable fields or otherwise processed by the database. |
Standardized concepts assigned to indexed records. |
| Handles synonyms automatically |
Not necessarily; database mapping features vary. |
Can bring differently worded records together under a common heading. |
| Useful for new terminology |
Yes, particularly when the terminology appears in recent records. |
Potentially less immediate because vocabularies and indexing must accommodate the new concept. |
| Available in every database |
Keyword or text searching is widely available. |
No. Subject indexing systems and their capabilities vary by database. |
| Portable across databases |
Concept terms can often be adapted, although syntax and searchable fields differ. |
Often requires checking the vocabulary used by each database. |
| Main risk when used alone |
Missing relevant records that use unexpected terminology. |
Missing records not yet indexed, not covered by the heading, or expressed through concepts the vocabulary does not capture adequately. |
Are Author Keywords the Same as Subject Headings?
No. This is a common source of confusion because a database record may display both.
Author keywords are terms supplied by authors or publishers to describe an article. Subject headings belong to a standardized indexing system and are assigned or mapped according to that system's procedures. They may overlap, but they have different origins and functions.
PubMed illustrates the distinction clearly. Its documentation identifies MeSH Terms as the NLM controlled vocabulary used to describe the subjects of MEDLINE articles, while its Other Term field can contain non-MeSH subject terms, including author keywords supplied by organizations other than NLM.
Does a Keyword Search Always Search Exactly What You Typed?
Not necessarily. Modern databases may process your query before executing it. PubMed, for example, uses Automatic Term Mapping for many untagged searches. It can map entered terms to MeSH terms and other equivalents, while also searching textual fields. As a result, typing an ordinary term into PubMed may already produce a combination of controlled-vocabulary and free-text retrieval.
This is an important reminder that the conceptual distinction between keywords and subject headings is clearer than the behavior of every search box. A database may map, expand, interpret, or transform what you type. When precision matters, inspect the database's search details, field definitions, thesaurus, or help documentation rather than assuming that the visible query is the complete search that ran.
Why Would You Use Both?
Keywords and subject headings compensate for different weaknesses.
Keywords can capture current language, unusual terminology, phrases in titles and abstracts, and records that lack completed subject indexing. Subject headings can retrieve conceptually related records despite variation in authors' wording. Combining the two can therefore improve retrieval when comprehensiveness matters.
This is especially relevant in systematic or otherwise rigorous evidence searches, but the principle is useful more broadly. The goal is not to make every search as long as possible. It is to represent each important concept adequately without adding terms that contribute little useful retrieval.
04 · A Practical Example
How Keywords and Subject Headings Work Together in a Real Search
Hypothetical Example
Searching for research on heart attacks
Imagine that you are searching PubMed for research about rehabilitation after a heart attack. You know the everyday phrase heart attack, but you also discover that the relevant MeSH descriptor is Myocardial Infarction.
Concept You want literature concerning heart attacks.
Keyword language You identify expressions that authors may use, such as heart attack and myocardial infarction.
Subject heading You check the MeSH database and identify Myocardial Infarction as the controlled-vocabulary descriptor.
Combined approach You search the relevant MeSH heading together with appropriate free-text variants for the same concept, normally connecting alternatives with OR.
Result The subject heading helps retrieve records consistently indexed for the concept, while the free-text terms provide additional routes to records through the language appearing in their searchable fields.
The important lesson is not the particular medical example. The same reasoning applies whenever a database provides a subject vocabulary. First identify the concept. Then ask two different questions: What words might authors use for this? and How does this database index this concept?
Those questions produce complementary search terms rather than competing ones.
07 · A Quick Checklist
Before Finalizing Your Search Terms, Check Both Routes to the Literature
Before finalizing your search, check:
Identify the major concepts in your research question before generating search terms.
List the natural-language terms, synonyms, and alternative expressions authors may use for each concept.
Check whether the database provides a thesaurus, subject headings, descriptors, or another controlled vocabulary.
Look up the authorized subject heading rather than assuming that your preferred keyword is also the database's preferred term.
Inspect relevant records to see which subject headings and author terminology appear repeatedly.
For important concepts, consider combining appropriate subject headings with free-text alternatives when the database supports both.
Check whether broader or narrower subject headings are being included automatically.
When moving to another database, verify its vocabulary and search syntax rather than copying subject headings unchanged.
