{"id":2349,"date":"2026-05-14T07:06:50","date_gmt":"2026-05-14T15:06:50","guid":{"rendered":"https:\/\/backseatlinguist.com\/blog\/?p=2349"},"modified":"2026-05-14T07:06:50","modified_gmt":"2026-05-14T15:06:50","slug":"does-memorizing-chunks-of-language-make-you-more-fluent","status":"publish","type":"post","link":"https:\/\/backseatlinguist.com\/blog\/does-memorizing-chunks-of-language-make-you-more-fluent\/","title":{"rendered":"Does memorizing &#8220;chunks of language&#8221; make you more fluent?"},"content":{"rendered":"<blockquote data-pm-slice=\"2 1 []\"><p>Boers, F., Eyckmans, J., Kappel, J., Stengers, H., &amp; Demecheleer, M. (2006). Formulaic sequences and perceived oral proficiency: Putting a lexical approach to the test. <em>Language Teaching Research, 10<\/em>(3), 245\u2013261. https:\/\/doi.org\/10.1191\/1362168806lr195oa<\/p><\/blockquote>\n<p>A few years ago, I went through the weaknesses of the arguments for <a href=\"https:\/\/backseatlinguist.substack.com\/p\/interviews-with-james-stubbs-on-future-multilingual\" target=\"_blank\" rel=\"noopener noreferrer nofollow\">rote memorization of vocabulary<\/a>. I recently stumbled upon a slight twist on this practice: memorizing \u201cprefabricated sequences\u201d of language rather than just individual words.<\/p>\n<p>The category of \u201cprefabricated language\u201d or \u201cchunks\u201d is rather broad in the literature. It includes things such as:<\/p>\n<ul>\n<li>\u201cIt makes no sense.\u201d<\/li>\n<li>\u201cMood swings.\u201d<\/li>\n<li>\u201cCouldn\u2019t handle the pressure.\u201d<\/li>\n<li>\u201cBring home the bacon.\u201d<\/li>\n<\/ul>\n<p>Common idioms, phrases, and collocations of various sorts all get thrown into this big pot of \u201csequences.\u201d<\/p>\n<p>Some have claimed that by memorizing these \u201cchunks of language,\u201d a second-language learner can become fluent in a much shorter amount of time than through other methods (such as getting lots of comprehensible input). This is in part because, the theory goes, these chunks of language tend to get processed by the brain more quickly than they would if broken down into individual words, and thus are less taxing on your working memory. Memorization will thus provide you with a shortcut to fluency.<\/p>\n<p>A study by Frank Boers and colleagues published in 2006 has been cited in support of this method. A closer look at Boers et al.\u2019s study, however, shows that there\u2019s no comfort provided whatsoever in their data for advocates of the chunk memorization.<\/p>\n<hr \/>\n<p>The researchers looked at two intact classes of college students majoring in English in Belgium. These were all upper-intermediate to advanced-level students.<\/p>\n<p>One group of students (n = 17) was taught using Michael Lewis\u2019s \u201cLexical Approach,\u201d which encourages learners to \u201cnotice\u201d (and use) these prefabricated sequences. The other group of students (n = 15) was given business-as-usual instruction that included grammar study. (So like a lot of intervention studies in L2 research, it was one bad idea vs. an even worse idea.) A total of 22 hours of class time was devoted to the experiment over a period of one academic year.<\/p>\n<p>The paper is rather vague about the details of what went on in the classroom. But we know that students were <em>not<\/em> asked to memorize the phrases, and there was no set list of prefabricated sequences to be covered in class. The noticing appears to have taken place (prefabricated sequence alert!) catch-as-catch-can in whatever regular reading or listening input was provided in class.<\/p>\n<p>At the end of the school year, both groups took a two-part oral test. One part of the exam was a spontaneous conversation about an \u201cunprepared but familiar topic, such as the student\u2019s traveling experience\u201d (p. 252). That lasted about 7 minutes.<\/p>\n<p>The other part, also 7 minutes, was a conversation about a short article \u201crelated to one of the themes in the course.\u201d According to the researchers, \u201cthe students were given 10 minutes to read the text and reflect on it before the actual interview started\u201d (p. 252).<\/p>\n<p>The conversations were recorded and judged blindly by two English as a foreign language teachers (one \u201cbilingual\u201d English speaker, one non-native speaker).<\/p>\n<p>The researchers reported that students in the experimental group were judged to be significantly more fluent than students in the control group. They also reported that the use of prefabricated sequences was positively correlated with these fluency scores.<\/p>\n<hr \/>\n<p>A few things to note about this study:<\/p>\n<p>1. All the students were upper-intermediate and advanced-level speakers of English, <em>not<\/em> beginning students. Many of the advocates of this chunk memorization approach (and rote memorization in general) sell it as a shortcut for beginning and low-intermediate students, but that\u2019s not a hypothesis tested in this study.<\/p>\n<p>2. Students didn\u2019t actually memorize any phrases(!), and there was no examination of how successful this noticing was, other than the final oral exam. Worse, we don\u2019t even know whether any of the phrases that were \u201cnoticed\u201d by the subjects in the experimental condition were used in their final exam. Again, this provides no support for any sort of rote memorization recommendation. It\u2019s simply not something the study looked at.<\/p>\n<p>3. The two parts of the oral proficiency test were very different, but their results were reported and analyzed together. One part of the exam was a true \u201cfree\u201d elicitation task, which is probably the best, strictest measure of oral fluency. Students had no time to prepare what they were going to say. The other half of the exam, as noted above, consisted of students reading a passage and being given 10 minutes to prepare for the subsequent conversation.<\/p>\n<p>This is where (alert!) the wheels come off this study.<\/p>\n<p>In their discussion, the researchers noted that \u201cthe experimental students outperformed their control peers mainly in the first half of the interview,\u201d which was about the prepared news article. They go on to say that \u201cabout 1\/3 of the formulaic sequences used in the interview by the average experimental student <em>were phrases encountered in that text<\/em>\u201d (page 256, emphasis added).<\/p>\n<p>In addition, there was no significant difference between the two groups in the use of prefabricated sequences of any sort in the free elicitation task (remember that the \u201cfluency\u201d perception was significantly correlated with the use of those sequences).<\/p>\n<p>In other words, advanced EFL students who had spent a year being taught to notice and use formulaic sequences (which, I have no doubt, made them <em>sound<\/em> more fluent than they were) did better than the control students <strong>only <\/strong>when they were given time to prepare and then regurgitate some formulaic phrases, many of which were from the article itself.<\/p>\n<p>The researchers noted, and I agree, that this was an excellent test-taking strategy by these clever students, sort of an English exam party trick.<\/p>\n<p>What it definitely isn\u2019t is evidence that noticing prefabricated sequences, much less memorizing them, is somehow a shortcut to second-language oral fluency.<\/p>\n<p>Boers et al.\u2019s data, then, say the exact <em>opposite<\/em> of what rote memorization advocates are claiming. In a free elicitation task, students who spent <strong>an entire academic year<\/strong> noticing these prefabricated sequences <strong>did no better <\/strong>than students who didn\u2019t go through the noticing training in terms of their oral fluency.<\/p>\n<p>This conclusion is very consistent with all the other research we have on so-called \u201cinstructed second language learning.\u201d When students have the opportunity to monitor, practice, or rehearse their performance (like students who study grammar rules and then take a grammar test), explicit instruction seems to work wonderfully. When those same students are taken out of that controlled setting and put into something approximating a real-world one, the effects of the instruction fade or disappear entirely.<\/p>\n\n","protected":false},"excerpt":{"rendered":"<p>Boers, F., Eyckmans, J., Kappel, J., Stengers, H., &amp; Demecheleer, M. (2006). Formulaic sequences and perceived oral proficiency: Putting a lexical approach to the test. Language Teaching Research, 10(3), 245\u2013261. https:\/\/doi.org\/10.1191\/1362168806lr195oa A few years ago, I went through the weaknesses of the arguments for rote memorization of vocabulary. I recently stumbled upon a slight twist [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":[],"categories":[8,9],"tags":[],"_links":{"self":[{"href":"https:\/\/backseatlinguist.com\/blog\/wp-json\/wp\/v2\/posts\/2349"}],"collection":[{"href":"https:\/\/backseatlinguist.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/backseatlinguist.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/backseatlinguist.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/backseatlinguist.com\/blog\/wp-json\/wp\/v2\/comments?post=2349"}],"version-history":[{"count":1,"href":"https:\/\/backseatlinguist.com\/blog\/wp-json\/wp\/v2\/posts\/2349\/revisions"}],"predecessor-version":[{"id":2350,"href":"https:\/\/backseatlinguist.com\/blog\/wp-json\/wp\/v2\/posts\/2349\/revisions\/2350"}],"wp:attachment":[{"href":"https:\/\/backseatlinguist.com\/blog\/wp-json\/wp\/v2\/media?parent=2349"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/backseatlinguist.com\/blog\/wp-json\/wp\/v2\/categories?post=2349"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/backseatlinguist.com\/blog\/wp-json\/wp\/v2\/tags?post=2349"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}