Semantic versus thematic clustering of vocabulary

This is a snippet from our forthcoming book about memory. This is from a chapter about remembering vocabulary.

 

Researchers have sought more efficient ways of learning vocabulary from lists. One popular comparison has been made between semantic and thematic vocabulary clustering types. Semantic clusters provide students with groups of words that are related by their meanings. For example, parts of the body, such as eye, head, ear and mouth. The argument for semantic clusters is appealing. Firstly, the similarity between the words should ease the learning task and secondly, the student should become aware of slight distinctions between the related words. In addition, most of us have been used to learning and teaching words in this way.

Nation (2001) argues that:

1.      - It requires less effort to learn words in a set.

2.      - It is easier to retrieve related words from memory.

3.      - It helps learners see how knowledge can be organised.

4.      - It reflects the way such information is stored in the brain (so-called semantic fields).

5.   - It makes the meaning of words clearer by helping students to see how they relate to and may be differentiated from other words in the set.

But the downside of teaching words in this way is that words of similar meaning may cause interference effects in memory. The closer two words are in meaning or association (including synonyms and antonyms), the greater the risk of interference and forgetting.

On the other hand, thematic clusters refer to the arrangement of a group of words that belong to a specific knowledge schema. The advantage is thought to be that memory is activated more powerfully when words are related to lived experience or episodes (knowledge schemas). So if you teach a group of words in the context of a lived experience the words should be easier to recall later. Tinkham (1997) suggested that arranging words by general theme in this way can limit the effects of interference between similar words. An example of a thematic cluster would be sweat-shirt, changing room, tries on, wool and salesperson.

So what is the evidence? On the whole, researchers now favour thematic word sets to semantic. Some studies report that semantic grouping is actually worse than presenting lists of totally unrelated words. So if your textbook presents words as semantic clusters you should at least question the validity of this approach, which may just stem from tradition. As Dronjic (2019) points out, thematic clustering is better on the whole than semantic clustering and better also than just listing words randomly.

Does this mean you should stop playing Simon Says to teach parts of the body? Not at all. Don’t forget the importance of motivation, distinctiveness and gesture in forming memories! In any case, all researchers agree that learning from lists, although apparently efficient, is a very small part of what learning vocabulary is all about.