SPEAK works on computers with RiscOS 3.0 or later although those with ARM2 processors will need to run with lowresolution graphics. Text to Speech Module. This produces synthesised speech on RiscOS computers. It sounds more natural, is easier to listen to, and has more accurate pronunciation, than previous Acorn based speech synthesis systems. It speaks British English not American and talks with an English accent which is hopefully neither overly Southern or Northern. The version 2 Speech Module now provides different voices and accents. You can call it from your own programs by passing it an English text string either as a * command, or by an SWI call. The speech module and its data files total about 250 kbytes of memory, which is far less than most PC systems. The speech module applies suitable intonation and cadence over each clause, taking note of punctuation such as commas, brackets, quotes, and question marks to vary this. Speak Talking Text Editor.Drop text files into it or importexport text from other applications. It can speak text continuously at various speeds, or a word or a clause at a time. It provides a number of different talkasyoutype options, and you can click on any word to hear it spoken. For ease of use, it has a button bar at the top of the window. Text can be edited in a similar way to in Edit while it is being spoken, making it ideal for proofreading. The cursor can follow the text as it speaks, and the window scrolls automatically. The speaking can be easily paused and continued at any point by simply pressing the escape key. This is not a full wordprocessor, but you can use it to prepare text for a word processor, or to speak text from one. Or you can use it by itself if you dont need text layout features.

Website: http://home.clara.net/jsd/speak.html