Joseph Friedrich Freiherr zu Racknitz claimed to have contructed a replica of the Turk in order to ascertain its secrets. He published his findings in 1789 in a book entitled Über den Schachspieler des Herrn von Kempelen und dessen Nachbildung.*
In English this means: About the Chessplayer of Mr. von Kempelen and its Replica.
While Racknitz put forth a credible explanation about how the Turk operated mechanically, he was unable to determine how the operator of the machine, unless he were a dwarf, concealed himself so effectively.
Racknitz' book is not only quite rare with copies difficult to locate, it's also written in German, of course, and in that fanciful, difficult-to-read script common to older German books. The information I found is all second and third hand, but adequate.
* not to be confused with Carl Friedrich Hindenburg's Über den Schachspieler des Herrn von Kempelen, nebst einer Abbildung und Beschreibung seiner Sprachmachine.
Joseph Friedrich Freiherr zu Racknitz was born in 1744 and died in 1818. He was a public official in Dresden (where his book was published).
Georg Gläser and Ernst Strouhal form the University of Applied Arts Vienna wrote a paper "Kemplen's chess playing pseuodo-automaton and Racknitz' explanation of its controls 1789"
According to their paper "Racknitz was eventually to become the first to construct two intricate replicas of Kempelen's automaton which enabled him to describe the controls of the "Turk" in detail."
A most interesting passage from their paper is "Details on the presentation and some explanations"
"Before we can discuss the questions of the controls, it is necessary
to accurately describe the automaton and the way it was presented by
Presenting himself to the audience as a serious mechanic, Kempelen
started his performance by announcing that the following was a
deception, a mechanical illusion. The automaton consisted of a life
size puppet in Turkish traditional dress, sitting cross-legged against the
back of a wooden chest. The chest (approximately 150 x 68 x 117 cm)
was mounted on rollers to prevent any interference from ropes or wires
under or from behind the stage. On top of the chest was a counter-
sunken chess board.
The front of the chest displayed three doors; each separated into
two sections. Under the doors, a drawer contended the full length of
the chest. Before beginning, Kempelen opened both sections, one
after the other. In the left compartment there was a confusion of cogs,
rollers and levers. The right side was practically empty. It contained
two quadrants and a cable winch at the top, as well as a pillow. He
also opened the lower drawer, which held the chess pieces. One after
the other he shone the light of a candle into each section. The
movement of the candle was visible behind the chest, which excluded
the possibility of a trickwith mirrors. Finally the puppet would be
shown sitting at the back of the chest. [In this section of the chest
there was also a smaller chest. At the beginning of the presentation
this was displayed several metres away. The smaller box had no other
function than to create confusion. Some observers suspected it of
containing a magnet which steeked the Turk.]
After the presentation, a volunteer was invited to participate in a
game. The Turk always took the first move and player left-handed. It
announced "check" by nodding its head three times. After a dozen
moves or so, the machine was wound up, apparently to give the player
a chance to clear his throat under the cover of the noise.
After the game, the Turk answered questions from the audience by
pointing at the appropriate golden letters on a board. Afterwards,
Kempelen himself was available for questioning. [The fact that the
Turk not only played chess but could obviously hear and understand was
of particular interest to the observers. For the German translators,
this was final proof that the Turk was not an autonomous automaton.]
From the first performances at the beginning of the 1770's the
interest was predominantly in the Turk's functioning autonomously,
which was considered to be entirely possible. Typical of this are Louis
Dutens letters * of 1771 and the first reposts from Karl Gottlieb von
Windisch in 1773. Windisch was a journalist and senator who later
became mayor of Pressburg. He fully represented the Enlightenment
and like Kempelen himself, was a freemason and had strong bonds to
the Kempelen family.
Because of this, Windisch's letters about Kempelen's chess player
should be seen as advertising tracts, however, changes are obvervable
in them between 1773 and 1783, without fear of ruining his reputation
as a man of the Enlightenment, Windisch gave the first critics who
suspected the presence of a player in the machine this answer:
'I have carefully inspected the table and the machine several times and
I can assure you with all confidence that there remains not the slightest
grounds to such a suspicion.' "
* Louis Dutens was an observer of the Turk who wrote several letters published in Le Mercure du France. He attempted to trick the automaton "by giving the Queen the move of a Knight, but my mechanic opponent was not to be so imposed upon; he took up my Queen and replaced her in the square from which I had moved her."
The authors cite two sources not readily available:
J.L. Boeckmann: Versuch einer Erklarung des von Hr. v. Kempelen
erfundenen mechanischenSchachspielers. Der erlauchten
Churfurstlich Mainzischen Academie der Wissenschaften zu
Erfurtehrerbietigst gewidmet. Carlsruhe 1789 (Posselt's Magazin
fur Aufklarung, Kehl 1785).
G. Bradford: The History and Analysis of the Supposed Automaton
Chess Player, of M. de Kempelen, now exhibiting in this Country,
by Mr. Maelzel. Boston 1826
After several pages of detailed descriptions, the authors conclude:
"The analyzation of Racknitz' description shows that he had in fact
probably built a machine that could fulfill all the tasks of Kempelen's
automation. The computer simulation, written with the program
system Open Geometry, works fine. Next we will try to rebuild
Racknitz' machine physically. Furthermore, we plan to write an
interactive computer simulation that can be run over the Internet
(Java-code!). The final goal is tobe able to play chess with a virtual
automaton. Therefore we plan to implement the chess engine of a
professional chess program."
Unfortunately, I couldn't find any indication that these plans were followed through.
Below are scans of six plates from Über den Schachspieler des Herrn von Kempelen und dessen Nachbildung.