Spelling Rules

Chapters 2 to 8 were about the symbols. A script alone does not tell you how to write a language: you also need the conventions for which letters spell which sounds, which way the lines run, and what to do when a language has to be written in a script that is not its own. This chapter covers direction, orthography, English spelling and romanisation.

Which Way Do the Lines Run?

From Wikipedia's Writing system article (first 100,000 of 154,316 characters read).

• Why left to right? The article gives one reason for the Greek pattern: the hand does not interfere with what is being written, because most people are right-handed.
• It can change. The article says horizontal left-to-right writing was widely adopted for Chinese-derived scripts only in the 20th century, due to Western influence.

Direction in Unicode

A computer storing mixed text keeps characters in reading order and gives each character a direction class. My own example, run and copied; the second part counts the ough-words in a sentence the English orthography article quotes.

L is left-to-right, R right-to-left (Hebrew), AL is Arabic letter (also right to left), EN is a European number and WS is white space. The Chinese character is classed L, which matches the modern left-to-right practice in the article. Digits stay left-to-right inside right-to-left text; that is my observation from the class, not a claim from the article.

Orthography: Shallow and Deep

From Wikipedia's Orthography article (whole answer as returned).

• Shallow. Spelling-to-sound correspondences are relatively simple and consistent. Spanish has a clear one-to-one correspondence between phonemes and graphemes, and Hawaiian uses only 13 letters.
• Deep. Correspondences are highly complex or inconsistent; the article's example is Navajo.
• Effect. Children learn to read faster in shallow orthographies like Spanish than in deep ones like English.
• Regulators. Some nations use language academies, such as the Académie Française and the Royal Spanish Academy. English has none.
• Reform. Noah Webster sought easily noticeable differences between American and British spelling (honor and honour). Modern Greek simplified its historical diacritics to a single accent.

Why English Spelling Is So Irregular

From Wikipedia's English orthography article (first 100,000 of 205,599 characters read).

• No one-to-one match. English has 24 to 27 consonant phonemes and 13 to 20 vowels but only 26 letters.
• One sound, many spellings. The sound /uː/ can be spelled in up to 24 different ways, and /iː/ in at least 18 to 21. For /ʃ/ the article lists sh, ch (French loans), s (sugar), t (ration), c (ocean), ss (tissue), sch (schedule) and x (Chinese loans).
• One spelling, many sounds. The letters ou can be pronounced in as many as nine different ways. The four-letter ough can be pronounced in at least ten different ways; the article cites Robert A. Heinlein's sentence "Though the tough cough and hiccough plough him through" to show it (the Python example above found six ough-words in it).
• Silent letters. Examples listed: l in talk and half, w in two and sword, gh in night and brought, and the silent e.
• Webster. He promoted more phonetic spellings in the United States, such as flavor, fiber, defense, analyze and catalog; the article says these forms already existed, but his dictionaries helped standardise them.

Romanisation

From Wikipedia's Romanization article (first 100,000 of 107,175 characters read). When a language needs writing in Latin letters for an outsider, the article distinguishes two aims:

• Pinyin has been used officially in China, adopted by much of the international community, and in Taiwan since 1 January 2009. The Revised Romanization of Korean replaced a modified McCune-Reischauer system used from 1984 to 2000.
• For computers, ease of creation and storage matters: Harvard-Kyoto aims at 7-bit ASCII, ITRANS was once prevalent on Usenet, and wāpuro rōmaji is "not strictly a system, but a collection of common practices" for typing Japanese.

Try It Yourself