On Taking English More Seriously

A critical examination of a segment of an article posted on CIOL (dot) org (dot) UK, titled “CIOL and English," in to answer the question that forms the title of the subheaded section “Why is English so difficult?” 

Why the [unnamed] author(s) choose(s) to formulate the subheading as a question rather than as a statement (“Why English is (so) difficult”) implies either resignation—unlikely on a site of the Chartered Institute of Linguists—or doubt-casting as to whether English actually should be considered difficult [to "master"]. Since it is never the less formulated as a question, I can already confidently say: “Because educational systems still take a Latin-based approach to a non-latinate language, preferring to parrot irrelevant terminology and analyses for describing English, rather than acknowledge and apportion more recent linguistics analyses of English and endeavour to ameliorate how it is being conveyed to learners worldwide.

I shall work piecemeal on the 243-word paragraph in question (shown in blue below), which begins by: 

[English] is this mish-mash of origins…”

—Really? Contemporary English has its roots in just three: Old English, Proto-Germanic and Germanic roots. Even if we included loanwords (though these are not its "roots"), these suggest inclusivity and flexibility—not complexity—of the language, while a "mishmash" may suggest appealing to the [current] mood of learners towards English, really it means a confusion and disapproval. (Wrong word choice, I would say, for starts.)

“…its irregularity and its status as a global lingua franca that make English such a challenging language to perfect.”

We need ask: 

i) What language doesn’t have its “irregularities”? (No natural language is without them.)

ii) What language is not a challenge? (Each has its idiosyncrasies, after all.) 

And ultimately, no, English is only designed and packaged to appear difficult when it is likely less complex than many other languages (if not equally complex). Its status is, as a result, possibly waning, as was proposed in the journal "The Linguist" (P. Harding-Esch, Vol/63 No/1; 2024; p. 6).

“English has a complex grammar system”

First of all, a language is never learned, neither by native speakers nor young non-native learners, through a “grammar system” in the way that a system is later imposed on us (throughout schooling). In any case, English’s complexity in terms of grammar—unlike other languages considerably more complex—derives from the fact that there are few identifying words and few aspects (e.g. have vs. be; suffixes ‘ing’ and ‘ed’) to express various communicative concepts, such as the pragmatic vs. irrealis, the latter of which involves backshiting* and simplification. Situational ellipsis, and contractions (elision), less discernible (not “discernable”) spelling standards (of which all are blameworthy for an insistence on standardisation for speed-readability), are phenomena that tend to be more problematic for non-native speakers. 

* Backshifting (which includes the use of apophony, e.g., will ➞ would, have ➞ had ) involves shifting a verb to “indicative” (base form) for future (unreal) contexts, or using the “past tense” to express “present” (or non-past) contexts. I have written extensively on that in my eBook “Insider English.” (Ask me about it.) 

“…with many irregular verbs and noun plurals.”

This typical illustration is an overgeneralisation since even regular verbs (ending in "-(e)d") have three different pronunciations, while so-called irregular verbs, which comprise only about a third of common English verbs, are largely an aspect apophony, and can be readily learned through sufficient and frequent oral exposure (something that is most often lacking in course books—instead you find an appendix of alphabetically arranged “irregular” verbs. It is rather the flexibility of English (e.g. the interchanging of verbs with nouns and adjectives) which can pose problems since learners are being taught to use dialectics. Next there is the art of concision, which often involves a preference for nounification (as that last sentence shows). 

As for “noun plurals”, one need get a sense of the notions of hypernymy and hyponyms as well as the flexibility (ditto) with which English natives sometimes apply nonconventional pluralisation despite hypernymy. Similarly, knowing that attributive nouns are not pluralised is a simple convention—not a complex one at that.

“There are many exceptions to rules,”

There are no exceptions until you apply [vague and general] “rules.” And the latter only came with the enterprising of English to generalise (what cannot be divorced from context and the cognitive) to its “clients” by those who (initially) understood very little of the linguistics of English. The misanalyses stuck merely out of a convenience of the commercialisation of English.   

“...making it difficult to master.”

Is that supposed to convey empathy toward non-native English speakers? 

What seems to be true is that those who follow the current traditional, parroted approaches are less likely to advance beyond “norms” set by academia; ironically, those who ignore these seem to speak more fluently and are more apt to go beyond intermediacy.

How do you explain that? Besides, if you cannot get passed mediocrity let alone intermediacy, mastery is inconsequential. By the time students graduate, the average level of English, if they learned it in school, rarely gets above "B1", and speakers will stay at that level unless they unlearn what they were "taught."

“English pronunciation is also a challenge as it has a wide range of sounds, and some of them are difficult to pronounce for non-native speakers.”

Which language is not a challenge to pronounce? How does one justify English as having a “wide variety of sounds” compared to other languages? Can we ignore the importance of intonation, let alone in those of Asian languages?

“The same words can be also be pronounced differently depending on the context and regional dialects.”

That is the easy part; often we will pronounce words differently explicitly to differentiate their function, which brings about another “complicated” factor (in detriment to how words are labelled) rather than consider this to be a compensatory factor as it should.

“English also has a large vocabulary with many words that have multiple meanings,”

Though English may have a large vocabulary, it is not necessary for “mastering” the language. Of English’s >600,000 count. Of the 170,000 listed as being contemporary, the average educated native will have 20,000 to 35,000, but use only ~ 2,000 to 3,000 words are used on a daily basis. Comparatively, Chinese master barely half of the 20,000 Chinese characters of the vast >80,000 characters.) So I don’t see how that argument holds. And Arabic has a much richer lexical range and has distinct words for each concept being expressed. (I welcome your comments on any discrepancies herein as I am not a specialist in languages other than English.) 

“…making learning and remembering new words challenging…” 

True learning comes from the ability to teach whatever it is one is examining. Rote memorisation (of “rules” and vocabulary) is a great encumbrance to effectively learning, as it means unnecessarily stuffing one’s brain “abstractfully”. Without building meaningful and personalised connections (synapses) between words, there is little productive communicative progress. 

“It is full of idiomatic expressions and colloquialisms …”

Any yet, discouraging learners from producing their own phrasemes is not particularly encouraged either, especially if we consider (what follows in the discourse): 

.”… these expressions can vary widely from region to region…”

If English is [indeed to remain] the global lingua franca, we should examine why speakers would hesitate generating their own creative “idioms”? Rather than highlight often outdated idiomatic phrases that will likely sound unnatural and out of context, learners will need to develop a sense of just how “idiosyncretic” (but perhaps also idiosyncratic) and, as I’ve already said, flexible English is by nature—especially, but not limited to, with regard to syntax. 

What we need ask ourselves is, What is actually stunting learners? What distortions and vagaries is mainstream syllabuses imparting about our language? Do we choose to cringe or tolerate agreement (the number of gender) even if it shouldn't impede intelligence? Why stress the logical over the irrational [human] nature of communication? Why impose worldwide standardisation when it is solely designed to keep learners in school—if not leaving school feeling in the dark about and disadvantaged for their lack of ability to actually communicate straightforwardly, rather than appear stunted, superficial or slack.

A more pressing question is: How is it that more recent linguistic findings are not instrumental in reforming how English is being conveyed?  And, What ripples do contemporary views on English make on social interactions globally-speaking? Does current linguistic mediocrity not rock the boat and spill into other fields and affect deportment?

“English spelling can be confusing, with many words that are pronounced differently than they are spelled, and many words that may be similar in spelling but have different meanings.” 

(A more wordy, convoluted verbalism for [the phenomenon of] homographs and homophones.) There are plenty of available tools already that expose learners to that phenomenon. (Actually, there are probably way too many posts on those sort of aspects of English, I doubt active learners can avoid studying.) Using a good e-dictionary with audio, not to mention the latest AI applications—which help to a degree, but are overpriced and overrated when it comes to sound language assistance.

Too few learners come to view a word in the way that one would learn, say, a Chinese character—a strategy I propose to my students: Visualise a word’s spelling as a complete unit, and combine that with the sound of that particular within a meaningful (and personalised) context; that is a most effective aid. Once again, it’s about synapsing and patterning. It’s about spotting norms, rather than memorising man-made rules. It’s equally about making up abstract, creative and even banal, imprints that don’t necessarily need to be “scientific”—or linguistic, even.

“There are many poles of first-language English-speaking from the UK to the Americas, Africa and Australisia [sic] all with their own character and characteristics.” 

Does that not reflect diversity in all things, English notwithstanding? 

In any case, all the more reason to reveal to learners which aspects are identical across all corners of the native English speaking world—and believe me, it’s not by studying “grammar rules” and “tenses” (of which there are only two) and the likes, that one will achieve greater fluency. Which brings us to…

“… 'loan words' … both welcome discoveries for people who share their language heritage, but more difficult to decypher [sic] for those who do not."

Discovering others’ cultural and linguistic richness involves person-to-person, voluntary contact and participation in meaningful discussions (as is more often done on social media platforms than in the classroom). My exposure to both tells me that distance learning is just as effective, if not more. Regrettably, involuntary migration and displacement of populations by a world run or influenced by military factions and corrupt political agendas, tainted by religious organisations and oligarchies, means that large populations are thrust into positions where they must be expected to take up English rather than learn each other's languages through cultivating genuine curiosity of our neighbours through a sound, unbiased education system that encourages inclusive discourse on all topics (not excluding religious beliefs), rather than promoting a single-story history and mainstream-media culture that leans increasingly toward a totalitarian system.

Alexander St-John

English Facilitator | CELTA Certified

4mo

NB: As I ran my article through the "immersive reader" AI-read option, it failed to pronounce "perfect" as a verb. So my remark about using AI "to a degree" is, to say the least, not unfounded.

Like
Reply

To view or add a comment, sign in

Insights from the community

Others also viewed

Explore topics