Emperor Taizu of the Khitan (AKA Liao) people introduced a script in 920 AD for his nomadic Mongolian nation. They had been using Chinese script, but the Chinese script was a poor fit for the Khitan language. Spoken Khitan had many syllables per word, unlike more monosyllabic spoken Chinese.
Taizu’s script (called “Large Seal”) was logographic and only a few symbols have been decipered. It clearly was influenced by Chinese, as there are a handful of symbols (particularly those used in dates and numbers) which are identical.
Only five years later, Taizu’s younger brother Diela came up with another script after a visit from the Uyghur ambassador. The ambassador showed Diela the Old Uyghur script, and something in the Old Uyghar script convinced Diela that he could make a better script than one the Emperor endorsed. This seems foolish to me, but I wasn’t there at the time, and Diela’s script (called “Small Seal”, duh) ended up in wide use amongst the Khitan people.
One innovation that Diela came up with was to show word boundaries by grouping syllable glyphs into word-blocks. Words had one to seven syllables, and they would all be crammed into one word-block, with an arrangement like Maya script: two wide, left-to-right, top-to-bottom. If there was an odd number, the last syllable would be horizontally centered, as shown in the block of glyphs above. The word blocks were then laid out linearly.
Small seal consisted of some logograms and some consisted of phonetic syllables, but only a fraction have been deciphered; the language died and there aren’t that many bilingual texts.
The Khitans used both scripts, but not in the same text (that we know of), and it isn’t clear why they used one sometimes and the other sometimes.
Links: Wikipedia, Ancient Scripts, Omniglot
Pingback: Jurchen Scripts — 1119 AD and 1138 AD, Northeast China | Glyph of the Day