Zi Yue English

You have memorised thirty thousand words and still stall on page one of a real English novel. Zi Yue English does not count the words you have studied: it tells you how much of a real English book you can read right now, and then opens it at the page you can already read. NOTE: THE APP'S INTERFACE IS IN SIMPLIFIED CHINESE. It is built for Chinese speakers reading English books, so the report, the notes and the stories are Chinese, and the only English is the book itself and the words it teaches. This description is in English so you can tell what you are buying; if you do not read Chinese, it is not for you. Measure first, then read 50 words, about 3 minutes, once. One word per screen, answer known or unknown. Six do not exist; claim those and the score is corrected (Meara & Buxton, 1987). Next launch returns to your report, dated. The report: one number, one interval, one walkable distance - How much you can read, computed word by word against the real vocabulary of 15 English novels - A confidence interval, so the number is not dressed up as precision it does not have - How far to go: how many words stand between you and 95%, and between you and 98% (Hu & Nation 2000; Nation 2006), with the most cost-effective words listed in order The main line: Chinese prose with English words inside A story written in Chinese, with exactly the words your report prescribed swapped into English. - The Chinese carries the sentence; the one English word is the only place you have to think - You guess from context; tapping confirms rather than looks up - Which words are swapped in is decided by your report, never by the author - Three levels cut from your own ranking; every swapped word appears at least three times - Open a word and it joins the review list - no quiz, and it comes back on its own The other direction: read the English original The report names the book you can read best; the button under it opens that book at chapter one. - The text is the original English, unchanged; the Chinese notes under each paragraph gloss the words in it you do not know yet, in order, so nothing is tapped while reading - How much is glossed follows your own measurement: three positions at the top, the middle one being your measured level - Swipe to turn a page: the pages run on through the whole novel, past chapter ends, and the page you stopped on is remembered One book, five ways to read it Nine of the fifteen works also ship as a full Chinese translation, with the words you do not know yet swapped back into English, and as a condensation whose length is about one tenth of the book's English word count (measured across the nine: 1/8 to 1/11). One book therefore sits on the shelf at three lengths, and six original stories in Chinese carry the same prescribed words. Translations and condensations are model-generated, not human translation; the English original is never edited. Why not "tap a word to look it up" Because you have already done that, and it is why you stopped at page 3. Looking a word up interrupts reading, and reading is the only thing that makes vocabulary stay. Read aloud, and offline The app picks a sentence from the top words in your report and you read it out; recognition runs on this device, the recording is never written to disk or uploaded. No network, no account, no data collected, no in-app purchases; six screens ask for no permission, and refusing the microphone costs them nothing. Nine screens. No streaks, no leaderboards, no notification prompts. Corpus 15 works first published 1813-1925, all public domain in China, the UK and the US at once: Pride and Prejudice, Frankenstein, Dracula, Moby-Dick, Walden and ten more. Chinese glosses and IPA from ECDICT (MIT licence). Names of people and places are not counted as words to learn. The corpus is classic literature, not contemporary speech: the number measures reading books like these, not an exam score.