James Hampton, pictured above, was a janitor in the District of Columbia who produced a remarkable piece of visionary art. His artwork was constructed of discarded furniture, burnt out light bulbs and assorted other garbage. Hampton's work was unknown until after his death in 1964.
Hampton also left behind a notebook filled with more than 100 pages written in a mysterious script. This script has been dubbed "Hamptonese". A partial page is pictured below.
To read more about James Hampton, his artwork, his life and his strange Hamptonese writing see
Ethan Le and I (OK, mostly Ethan) have been conducting research on Hamptonese. The first task was to transcribe Hampton's writing into a computer-friendly form. This proved to be a non-trivial and time-consuming task.
We first developed a key in which each Hamptonese symbol is transcribed as a few (1 to 3) alphanumeric characters. Of course, there is no significance to the character strings used to represent particular Hamptonese symbols---they were selected simply because they were easy for us to remember when transcribing the text. The key can be found here.
Using this key, we have transcribed the main body of Hampton's writings, which consists of precisely 100 distinct handwritten pages. Our complete transcription can be found (in plain text) here, where we have used "*" to indicate an undecipherable character and we have preceded any questionable characters by "#". Hampton occasionally intersperses English with Hamptonese. We've included any such English text in our transcription, enclosed within square brackets.
Scanned pages of Hamptonese are available (in jpeg format) here.
We have completed a basic analysis (symbol counts etc.) of the Hamptonese text. We have nearly completed a Hidden Markov Model (HMM) analysis of the text. This should reveal any underlying statistical properties of the text with a minimum of pain. For example, this will reveal whether the text is a simple substitution for English (or another language) while potentially revealing other information about the structure of Hamptonese. This page will be updated with our research results in the near future.
If you want to know more about HMMs, I've written a revealing introduction to hidden markov models.