Working with the General Service List of English Words in Practice
The General Service List Of English Words is a curated vocabulary list of the 2,000 most frequently used words in English. It was originally compiled by Edward and Margaret West in 1936 and revised in 1956. The list covers roughly 65-70% of most general English texts. That is useful if you are teaching beginners or building learning materials, but it comes with some real limitations that nobody talks about enough. You can find clean versions of the list on several academic sites. The most reliable source I use is the Corpus of Contemporary American English (COCA) word lists page, where the original West frequency data is archived. Project Gutenberg also has plain-text versions available free. I keep a CSV file locally and map it against my own word frequency data when I need something current. The raw 1936 list is accurate for its time period but outdated for modern English. The 1956 revision improved things somewhat. Neither captures the shift toward digital and technical language that happened after the 1990s. What the list actually contains matters more than where you get it. The original list includes words like "bicycle," "telephone," and "automobile" at higher frequencies than you would expect today. Those words appear often in the source material from the 1920s and 30s but carry much less weight in current usage. If you build a curriculum strictly around the 1956 revision without checking against modern corpora, your students will learn words that are either archaic or niche. That is not a small problem.
How to actually use this list instead of just downloading it
I run a filter pass over the raw list before using it in any material. I compare the 2,000 words against COCA frequency rankings from the 2000s to the 2020s. Words that dropped out of the top 2,000 in modern usage get flagged. Words that entered the top 2,000 recently get added back. This usually takes about twenty minutes if you have the data on hand. The result is a hybrid list that keeps the pedagogical structure of the original while covering current usage. Here is an example of a specific edge case that costs people a lot of time if they miss it. I was preparing a reading comprehension set for intermediate ESL learners and used the unmodified GSL. One of my texts included the word "subway" because it appeared frequently in older sources on the list. In American English, "subway" is standard. In British English teaching contexts, which made up about forty percent of my learners, the word caused confusion because their exposure is to "underground" or "tube." The word was on the list, so I should have known better than to drop it in without context. I replaced it with "railway station," which is on both the original list and more current frequency data. This took about five minutes and saved a follow-up session that would have been entirely spent clarifying vocabulary. The fix is straightforward: always check the target dialect before selecting vocabulary from the GSL. The list was built primarily from American sources but was adopted worldwide. British, Australian, and other English variants have different frequency profiles for common words. "Holiday" versus "vacation" is the most obvious example. "Lorry" versus "truck." These are not rare words. They are just not on the same frequency curve across dialects.
Limitations you need to accept
The General Service List covers high-frequency words, not high-frequency grammar. Knowing the 2,000 words does not mean a learner can construct a grammatically correct sentence. The list treats content words and function words as equally important, but function words like "the," "of," "and," "is" carry disproportionate structural weight. A student who knows the vocabulary from the list but has not mastered grammatical patterns will still struggle with basic communication. This is not a flaw in the list itself. It is a flaw in how people try to use it as a standalone solution. Another limitation is register. The GSL draws heavily from general prose, fiction, and newspapers. It underrepresents academic vocabulary, which is where most advanced learners eventually hit a wall. If you are preparing someone for university-level English, the GSL will get them to B1 or low B2 at best. You need the Academic Word List, the New General Service List, or a domain-specific frequency list once they move past the basics. There is also the problem of polysemy. The list counts each word form, not each sense. "Run" appears once on the list. It has over sixty distinct meanings. Teaching a learner that "run" is a vocabulary item they have "mastered" if they know one definition is misleading. I handle this by tracking word senses separately in my materials and flagging high-polysemy words during instruction. It adds about ten minutes per word to lesson preparation, but it prevents the confusion that comes later.
Get the Full Details
Practical workflow for teachers and material creators
Start with the 1956 revision. Map it against COCA or BNC frequency data for your target dialect. Remove words that have dropped below the frequency threshold. Add words that have risen. Check for dialect-specific gaps. Cross-reference with the New General Service List if you need a more modern equivalent. The New General Service List, published by Hofmann and Schmitt in 2012, is worth looking at even if you do not use it as your primary list. It offers a cleaner frequency methodology based on larger, more balanced corpora. If you are building flashcards or spaced repetition decks, do not include every word from the GSL at once. A typical intermediate course covers about 1,500 to 1,800 of the 2,000 words effectively. The remaining words are either highly infrequent in practice or duplicate entries that appear under different forms. Trying to teach all 2,000 in one pass leads to burnout. Space the list across multiple courses or semesters. I usually allocate about eighty words per week for a steady intermediate course. That pace is sustainable and allows for review without feeling repetitive. For curriculum mapping, group the words by semantic field rather than by frequency order. The original list is arranged purely by frequency, which makes it awkward to organize into lesson topics. A thematic approach to vocabulary instruction tends to stick better with learners. "Food and drink," "transportation," "household items" — these categories pull words from across the frequency spectrum and make the material more coherent. I spend about fifteen minutes per topic creating these thematic groups from the GSL data, and it pays off in retention.
The New General Service List as a practical alternative
If you are starting fresh and do not need historical continuity, the New General Service List is a reasonable substitute. It uses more modern corpora and a clearer methodology. The frequency ordering is more accurate for contemporary English. Hofmann and Schmitt's work is openly available and widely cited. Some educators skip the original GSL entirely and go straight to the New General Service List. That is a defensible choice. The original list still has value as a teaching artifact and historical reference point, but its frequency data is simply outdated for modern usage. One thing the New General Service List handles better is the inclusion of less common inflected forms. The original list sometimes lists a base form and its variations inconsistently. This creates ambiguity when you are generating exercises or quizzes. I run all my lists through a morphological analyzer before creating materials. It takes about ten minutes and eliminates problems with incorrect word form generation. Ultimately, the General Service List Of English Words is a starting point, not a final answer. It is useful if you treat it as raw data that needs cleaning and adaptation for your specific context. It is wasteful if you treat it as a complete vocabulary solution. The difference between those two approaches shows up clearly in student outcomes within the first semester.
