Thursday, February 15, 2007

Masterbook of Data from Online Harvard Khipu Database in a single XLS file


I created a "Masterbook" containing the data for all the Khipus in the online Harvard Khipu Database (as of 15 Feb 2007) in a single XLS file.

For your further use/study, I placed this "Masterbook" here.

Dennis (moderator)

Tuesday, February 13, 2007

An Asian Pacific Connection to Andean Khipus??

A fascinating suggestion made by Jean Jacques Quiquater's PDF slide presentation at MIT's recent symposium on Khipus is that the use of knotted chords as mneumonic devices appears to have been common throughout the Asian Pacific world.

He notes that references by European explorers and missionaries exist to the use of knotted chords for mneumonic purposes by native peoples across the Polynesian world - in Hawaii, Samoa, the Caroline Islands, by the Maoris of New Zealand and even in Okinawa!

Dr. Jean Jacques Quiquater's presentation along with internet references is available at:

The pertainant slides to the point discussed here can be found between pages 90-107.How fascinating?

Does anybody know anything more about the chord writing of the Asian Pacific peoples?


PDF Slide presentation from Recent MIT Symposium on Khipus

A wonderful PDF slide presentation from a recent MIT Symposium on Khipus can be found at:


Natural (Undied) Multicolored Cotton of the Andes

Today we often think that cotton comes naturally only in white or of-white(tan/beige) shades.

However, in the pre-Colombian Andes, and even to this day, cotton grown in the Andean region of South America can come in a multitude of undied natural colors.

Several articles and webpages about the natural multicolored cotton of the Andes can be found at:


Pre-Inca Era Khipus (!!)

Khipus are often associated with Incas. However, it may be of interest, that Khipus from pre-Inca times have been recovered.

Several articles (in Spanish) about pre-Inca khipus can be found in the anthology by C. Mackey, et al, eds, Quipu y Yupana: Coleccion de Escritos (Lima: Consejo Nacional de Ciencia y Tecnologia, 1990).

And several articles about a recent finding of pre-Inca era khipu (perhaps as much as 5000 years old) at the Caral archiological site in Peru can be found at:

Note that a characteristic of these pre-Inca era khipus appears to be that the khipus' pendant chords are wound around a (hollow?) center and thus have a tubular appearance. In spanish khipus of this type are called "quipus de canutos."


Sunday, February 11, 2007

The Knots Present in the Khipus from Leymebamba (UR001-UR022)


William Burns Glynn has arguably gone for a "home-run" with his 10 consonant theory for Inca writting, be it by means of Tocapu signs stitched onto Inca ceremonial garments, or by means of khipu themselves.

At the other end of the spectrum, would seem to be Harvard Professor Gary Urton, who sought to approach the task of Khipu decipherment in a most general and systematic way, tabulating the various characteristics of both the Khipu pendant and subsidiary chords themselves (color, ply [S/Z]). attachment to the main chord, and end) as well as properties the knots found on these chords (type, number of loops, and directionailty[S/Z]).

Equipped with such tabulated data, one can then look both for patterns, and the exceptions to patterns. The discernment of exceptions to the norm, can yield significant information. These exceptions can be assumed to be intentional, hence definitely encoding some kind of information.

Obviously, one would not know information what information each exception would encode. However, with further study, one could perhaps discern a second (deeper) level of pattern in those exceptions.

However, the first step is to produce good tabulated data.

A master databook of the Khipus originating from Leymebamba (UR001 - UR022 of the Harvard Data base) is given here.

I have previously noted that Khipu UR016 in the Leymebamba series appears to be unique in the organization of its constituent pendant chords. Normally, pendant chords in the khipus of Leymebamba series would be organized in rather small or tight groupings of 5-8 pendant chords. In the case of Khipu UR016, there are three groupings of pendant chords that are far larger (looser) than this. These groupings contain 101, 75 and 30 pendant chords respectively. Such looser groupings, suggest a looser, less formated type of data present (ie possibly narrative).

The relative uniqueness of Khipu UR016 can further be seen when comparing the frequency of the various types of knots present in the Khipus of the Leymebamba series.

Leyland Locke noted that khipus from the Inca-era exhibited three types knots which he called simple (S), looped (L) or figure-eight (E) style knots. Gary Urton noted further that these knots can be tied in two directions, that is, showing directionalities of either S or Z.

Thus there are six types of knots possible: SS, SZ, LS, LZ, ES or EZ.

A survey the frequency of the types of knots present in the Khipus in the Leymebamba series (UR001-UR022) is given here.

Such a survey yields the observation that normally all the S knots and L notes in a Khipu are oriented in the same direction (either S or Z). However, Khipu UR016 is an exception. First, the S and L knots are oriented in opposite directions (S knots are predomiantly oriented in the S direction, while almost all the L knots are oriented in the Z direction). Second, most S knots in UR016 are of an S direction, a insignificant number of S knots (nearly 20%) are oriented in the Z direction. When one considers that often a string of S knots present on a pendant chord from Khipu UR016 begins with an SZ knot followed by a string of SS knots, the presence of exceptional SZ knot at the head of the longer string of SS knots suggests that the expectional SZ knot encodes some special meaning.

A further survey of exceptional knot combinations present in the pendant and subsidiary chords of the Khipus of the Leymebamba series is given here.

Finally, a more detailed survey of the frequency of the types of knots present in the Leymebamba series (in this case taking into account number, that is, the number of single knots present on a pendant or subsidiary chord, or the number of loops present in the looped knots present on such chords) is given here as well.

The point of all making such tabulations is once again to seek to identify both patterns as well as exceptions to them present in the khipus from this (Leymebamba) series.