Home » Thoughts on Education » Ghoti = Fish (no it doesn’t)

Ghoti = Fish (no it doesn’t)

Ghoti spells Fish. This thing still shows up on social media and still makes me sigh heavily for two reasons:


It still gets used by literacy professionals as 'proof' that the English language is 'not phonetic'.

Of course, we're honest that English has a complex written code. There are a lot of ways to represent the sounds that we say. However - and this is a big deal for teen and adult literacy learners - the language is most certainly phonetic and the phonetic code is limited and learnable.

If you want to learn more about the patterns and frequency of phonemes and graphemes in English, get a copy of Greg Brooks' Dictionary of the British English Spelling System

According to Professor Brooks:

<gh> represents /f/ in draught, chough, cough, enough, laugh, rough, slough, sough, tough and trough.

<o> represents /i/ only in women

<ti> represents /sh/ only when followed by the letters <a, e o, u> He gives the examples: confidential, inertia, infectious, nation, quotient and Ignatius

Proof that ghoti cannot ever spell fish in English.


Erroneously telling struggling readers that ghoti spells fish is creating a barrier to reading when our job as educators is to tear down barriers as thoroughly and quickly as we can.

Saying to a vulnerable learner, 'Don't worry, you're right, this language is a mess and really really difficult' is actually saying, 'You're right, you'll probably never learn this'.

it is simply wrong to think that learning the English code is beyond struggling readers or that they can't read because the language is so complex. Rather:

  • Most teens and adults arrive at their first TRT lesson simply not knowing that the squiggles on the page represent the sounds that we say out loud.
  • They have personal sight vocabularies but don't know how to approach new words that matter in age-appropriate and life-appropriate text.
  • If they haven't completely given up on reading for meaning, they rely on skipping unfamiliar words, guessing (badly) from initial letter and shape and looking for unrelated words within words (like hen in then).
  • They have quite a lot of latent knowledge about the English code but need some way of unlocking it for reading and spelling.

During the first few hours of That Reading Thing they learn to read through long words from left to right and listen for a word they know. The goal is to get our students from reading words like fax and plug to words like conditioning and instructions in 5 hours or less. Most will be able to spell astonishment and establishment in hour 3. When they get into the Advanced code, they learn there are patterns (not rules) to how we spell various sounds and read various graphemes. They learn that English isn't random or chaotic and that they can approach unfamiliar words with a consistent plan which frees up their working memory to think about meaning.

Learning to decode using age-appropriate phonics is not a boring slog but a satisfying challenge.

14 Responses

  1. Pascal-Denis Lussier

    Although I sincerely appreciate your motivation for wanting to refute GBS’s amusing respelling of “fish”, I also initially sighed upon reading your take, which obscures his motivation and an indisputable reality re the current state of English spelling and pronunciation.
    You’re right that anyone claiming that his ‘ghoti’ proves that “English isn’t phonetic” is wildly misunderstanding what phonetics is or what it’s about—standard stuff when field-specific concepts and jargon are presented by, and at, the popular level—but, unfortunately, the manner in which you present ‘decoding’ is contextually not relevant, as for the goals and the examples you present; “accomplishment” may be a longer word, but all that’s wrong with English, which GBS’s ‘ghoti’ captures and illustrates, isn’t exemplified in what you present/discuss.
    The real issue: There’s a serious lack of fixed and logical paradigms, and no direct and stable correlations between spelling, environment, and prosody (accent, stress, and vowel length) and proper/accepted pronunciation, which a phonetic examination of English clearly illustrates.
    This is related to the fact that, due to its place and development, over 80% of English words are, in fact, borrowed from other, unrelated languages (such a Latin, French, and Greek) rather than having evolved and descended from a parent within the same language family. No other language presents such proportions.
    As such, English is plagued by historic forms that have nothing to do with the original pronunciations that gave us these spellings (such as knight, originally pronounced ka-nicth), as well as retaining old forms for words affected by ‘the great vowel shift’ of the middle ages, as well as forms corresponding to the spelling from whence it was borrowed, which deviates from exemplars / cases proper to English. All of this is extremely confusing, and counter-intuitive, for L2 learners, especially past a certain age, and depending on the rigidity of their L1.
    English is badly in need of a spelling reform, period.
    As a linguist and, therefore, a descriptivist, I see no problems with adopting logical forms like ‘nite’ for ‘night’, etc., which prescriptivists (grammarians) are determined to oppose.

  2. Tricia

    Hi Pascal-Denis, Thank you for taking the time to respond. First – YES! I have a degree in English Language (first UBC grad through that programme) so I understand what you’re talking about and I technically support spelling reform. However, that’s a top-down answer and sociolinguistically complex. Do you remember when the Canadian print industry removed the u’s to align with the Americans’ arguably simpler spelling to save ink or something and Canadians demanded they be put back? It was an economic decision that was reversed for reasons of national identity. People don’t necessarily want spelling to be simple as every spelling reformer has discovered.

    We start from the mundane truth that if an adult writes “I’m avalable for nite shifts” on a job application, they are less like to get the job than the person who writes ‘available’ and ‘night’ even if they’re better qualified. We work on tearing down barriers to writing and reading within the current system because currently there’s only one written English of power. Telling vulnerable learners the language is too complex for them by showing them examples like ghoti = fish is excluding them from that language of power. It’s as political as it is linguistic.

  3. Jameson

    Hi Tricia, Thank you for this The conversation between two professionals was certainly interesting and informative but as you explained the reality of applying these theories in the real world is never that simple; there are many other factors to consider

  4. Adrian

    As a non-professional, I enjoy such examples as Ghoti, it’s fun! I also love, and am often tripped up by our bewildering spelling scheme (skeem? 😁). As stated, English is a rich mix of languages, and all the better for it IMHO. When one traces the etymology of words you trace your history. The ‘dumbing’ down of spelling as we see across the pond, probably has an element of political 4th July basis and possibly (no disrespect meant) a certain level of education of language within the population, not to mention a large ‘foreign’ input.
    Let other countries that use versions of English (and there are many) do as they wish, but leave the mother-tongue alone. I see any attempt to rush the natural progression of language to be a stamping upon both history and sovereignty.
    Apart from vowel shifts we also have ‘standardization’ brought about by the printing press (ever read 14 century hand written text with multiple spellings of the same word on the same page?) And of course the loss of letters like the thorn.
    More power to your elbow in teaching ‘real’ English.

  5. Tricia

    Thanks for commenting, Jameson. You’re right, we can have all the theoretical discussions but, in the end, it’s one person at a time learning to read and write and that’s what has kept me going all these years.

  6. Nick

    Your title and first sentence implies a thesis that you don’t even begin to prove. If it’s your claim that ghoti does not spell fish then you have yet to make Amy argument at all to that end. If your claim however is that educators should take care not to confuse and demotivate students then well argued, but poorly titled.

  7. Tricia

    Thanks for commenting, Nick. I think the evidence presented under point 1. is compelling proof that ‘ghoti’ cannot spell fish. I truly wish I had the time to ponder my titles more carefully. Perhaps I’ll rephrase it as a question. What I do know is that I wouldn’t get nearly as many readers of this post if its title were ‘Don’t confuse your students’! People are interested in ‘ghoti’ and I hope they will get a copy of Greg Brooks’ dictionary to discover the code for themselves.

  8. Tricia

    Hi Adrian, Thanks for your comments. The ‘ghoti’ thing might be fun but it makes the language seem more bewildering than it actually is. Simplifying spelling is no more wrong than keeping spellings from history or even keeping spellings based on errors in the past, though I do struggle with simplified spellings that lose the morphology of a word. I’m thrilled that the language is alive and ever-changing but my main job is to teach, encourage and equip people who wrongly believe that they can never master it.

  9. marlene

    theoretically, ghoti spells fish. in practice, not really. i’m 14 and i think i have more brains than any harvard professor. nothing personal, most people can’t handle the ghoti supporter lifestyle.

  10. Tricia

    Hi Marlene (or marlene if you prefer),
    Thanks for your comment.
    Theoretically, fish can be spelled tnlllqp because language is created by people and can be changed by people, so I take your point.
    Sadly, we live in a world where only the very privileged like you and me can break the rules of English and get away with it.

  11. Sezen

    The point of the ghoti thing is to point out the relative lacking of grapheme-to-phoneme consistency. There’s probably no language on Earth that has a perfectly consistent grapheme-to-phoneme representation outside of conlangs given dialectic variability (Finnish is the one I can think that has EXTREMELY shallow orthography…) If those conlangs usage expanded, it’d probably result in similar variability. English is extremely orthographically opaque and I think most studies+linguistic and speech science scholars seem to agree.

    I’m unsure if you’re ever learned a different language, but the two I put a lot of time in were on opposite ends of orthographic depth. My first language is English, but my first language I learned was Dutch. Dutch has a similar orthographic depth to English, and despite me being a lot younger and having a good support system to learn, the morphology made reading and writing pretty complicated even coming from a language where such morphological-orthographical features are common. I achieved a point of near-fluency after a few years, but continued to have problems with certain representations. Later in adulthood (not having used the language consistently in 12ish years withstanding…) and being able to see things from a linguistic POV, I can largely see the guessing game and forced memorization.

    The other language I have found a lot of interest in was Hungarian, which, similar to it’s sister language Finnish, very shallow. I studied it for about a year and a half and I have yet to achieve a high level, nor have I actively studied it for 3 years. However, that shallow orthography for both reading, writing and listening made the ability to communicate and transmit my ideas so easy. To this day, if I hear a word and need to look it up, I can just listen and know immediately how to spell it, despite not using it for years nor achieving anywhere near as high of a level. It’s absolutely crazy how much it impacts the learning process in my opinion.

    Real world often skews from practice, I’m not going into teaching ESL or strictly improving literacy, I am going into SLP. if you have good luck in what you’re doing more power to you! However, with my language learning experience and my knowing of many ESL people, I have an incredible amount of respect for those whom command English at a high level as a second language. I believe it’s an upside-down pyramid structure, pretty easy to get into the basics of but has a lot of “isms” and complexities that make it very difficult. The antithesis would be Hungarian, which is intimidating and complex at the introductory/basic level but becomes easier, more consistent and predictable once you go on.

    I say this to say, I do not think it’s a bad thing to respect people’s intimidation of the language. If a learner has a lot of trouble with the orthographic inconsistency, I do not think it’s a bad thing to say “yea look at this silly little example, this language DOES have a lot of weird quirks” especially with students who grew up with more consistent and shallow languages. Moreover, while it can intimidate some and shy some away, I believe it makes the process of improvement very rewarding and empowering.

    I had a bad reaction to this article at first, but once I began to write this I started to see where you’re coming from. I just think that I disagree in terms of a board approach, and some students and learners may appreciate the recognition from their servicer that what they’re feeling is real.

  12. Tricia

    Hi Sezen, thanks so much for taking the time to write. I suspect we would agree on just about everything. However, That Reading Thing isn’t a tool for teaching English to speakers of other languages from the ground up. I wrote it originally for local English boys who had a big gap between what they could say and understand (as native English speakers) and what they could lift off a page. They had been trying to learn to read for years so the ‘ghoti’ example was just discouraging. It’s fine say that the language is quirky/complex/incredibly rich, but most teachers go straight for the ‘difficult’ bits and forget to teach how predictable much of the language is, especially for reading.

  13. Andrew

    Do you, by chance, work with students who have both dyslexia and ADHD and are trying to pursue a bachelor’s degree?

  14. Tricia

    Hi Andrew. Thanks for your comment. We train tutors rather than work with individuals. If you message me your location via the contact form, I’ll ask the TRT community if there is anyone in your area.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.