jTRACE
An easy-to-use, cross-platform, reimplementation of TRACE with a new
graphical user interface and many other features
jTRACE is a freely-available re-implentation of the TRACE model
of spoken word recognition. It is implemented in Java, and runs on any
computer that supports Java. Researchers can run simulations quickly
and easily, using code that has been validated against the original
TRACE code. Advanced programmers can extend the jTRACE model to
implement new behaviors.
jTRACE was created by the Magnuson Lab at the University of Connecticut with the
support of the National Institute
of Deafness and Other Communications Disorders.
Wait a minute, what's TRACE?
TRACE is a highly influential model of speech perception and spoken
word recognition, created by Jay McClelland and Jeff Elman (1986). The
original implementation of that model was used to run dozens of
simulations comparing TRACE's behavior with results from experimental
studies with human subjects. TRACE's behavior accounted for human
behavior in a number of important ways, and it is still frequently
cited as the canonical interactive-activation model of speech
perception and spoken word recognition. TRACE has proved remarkably
robust, accounting for results in paradigms that were developed 10
years after the model (e.g., fine-grained time course measures from
eye tracking).
Features, and version history
- jTRACE Screenshot:

- Underlying model is functionally equivalent to the original TRACE code.
- Graphical user interface is powerful and easy to use.
- Node activations are visualized as the simulation progresses.
- Graphical analyses of word and phoneme activations over time.
- Powerful scripting capabilities.
- Parameter and other files are saved in flexible XML format.
- Alpha.55 (current version)
- Phoneme panel for editing/adding phoneme representations
- 'Duration scalar' parameter for modifying the temporal extent of phonemes.
- Input panel for designing ambiguous and spliced input representations
- Lexicon validation; sorting on lexicon table.
Requirements
- Any computer with Java 2 1.4.2 or
later installed (most up-to-date computers will already have Java)
Downloads
The zip file below is for all operating systems.
There is no installer.
- To get started, unzip the contents of the file into
a new directory.
- On Windows run the "jtrace.bat" file.
- On Linux or Mac OS X, run the "jtrace.sh" file. (In Linux or OS X:
open a terminal window, get to the new directory where you have
unzipped jTRACE, and enter "./jtrace.sh" or "sh ./jtrace.sh"). See
the Release Notes.
NOTE: The following is a functional but not polished pre-release
(beta) version of the software. Significant interface and usability
bugs may remain.
- Download jTRACE
(v.a63, 12.8 mb, last updated 08/20/2009)
- Release notes
- This appendix to the user's manual distinguishes
between the analysis options available in the graph panel.
- We'd like to know who is using jTRACE and what we can do to
help (when we have time). Please send an email to
Jim Magnuson
to let him know you're using the system. We'll also keep
you updated on improvements to the software.
For advanced users
- If you are interested in obtaining the jTRACE source code,
please email ted.strauss@uconn.edu.
- People interested in validating jTRACE against the original
implementation may download our instrumented version of TRACE,
called cTRACE. Instructions are included. ctrace.zip
Documentation
Tutorial
Contact
For assistance running jTRACE, or technical questions, contact Ted Strauss. For
questions about psycholinguistic modeling with jTRACE, contact Jim Magnuson. For
questions or problems with this web page, contact Jim Magnuson.
Citation
If you use jTRACE in your research, please cite this article:
References
There are three foundational papers on TRACE. Jay McClelland
has made them all available on his publications page.
- McClelland and Elman (1986). The TRACE model of speech
perception. Cognitive Psychology, 18, 1-86. [This is the main
original paper about the familiar version of TRACE that maps
from subphonemic representations to words, which is
technically "TRACE II".]
- An overview of this paper is presented in chapter 15 of:
- McClelland, J. L., Rumelhart, D. E., and the PDP research
group. (1986). Parallel distributed processing:
Explorations in the microstructure of cognition. Volume
II. Cambridge, MA: MIT Press.
- Elman, J. L., & McClelland, J. L. (1986). Exploiting the lawful
variability in the speech wave. In J. S. Perkell and D. H. Klatt
(Eds.), Invariance and variability of speech
processes. Hillsdale, NJ: Lawrence Erlbaum Associates,
Inc. [This chapter describes TRACE I, which focused on mapping
from actual speech to phonemes. It was, however, abandoned. The
time may be right for re-examining the positive aspects of this
approach...]
- McClelland, J. L. (1991). Stochastic interactive processes and
the effect of context on perception. Cognitive Psychology, 23,
1-44. [In this paper, McClelland develops a stochastic
(nondeterministic) version of TRACE to answer criticisms of the
original model.]
Some other papers using TRACE:
-
Allopenna, P. D., Magnuson, J. S., and Tanenhaus,
M. K. (1998). Tracking the time course of spoken word
recognition using eye movements: Evidence for continuous
mapping models. Journal of Memory and Language, 38,
419-439.[The paper establishing the linking hypothesis
between eye movements in the visual world paradigm (VWP) and
response probabilities derived from TRACE.]
- Magnuson, J. S., Dahan,
D., & Tanenhaus, M. K. (2001). On the interpretation of
computational models: The case of TRACE. In J. S. Magnuson and
K.M. Crosswhite (Eds.), University of Rochester Working Papers
in the Language Sciences, 2 (1), 71 - 91. [The paper where we
tested 3 implementations of frequency for TRACE.]
-
Dahan, D., Magnuson, J. S., Tanenhaus, M. K., and Hogan,
E. M. (2001). Tracking the time course of subcategorical
mismatches: Evidence for lexical competition. Language and
Cognitive Processes, 16 (5/6), 507-534.[This paper
demonstrates that the VWP linking hypothesis can predict very
subtle patterns of eye movement behavior, including changes as a
function of whether or not competitors are present in the
display.]
- Dahan,
D., Magnuson, J. S., and Tanenhaus, M. K. (2001). Time course
of frequency effects in spoken-word recognition: Evidence from
eye movements. Cognitive Psychology, 42,
317-367.[The paper where we tested 3 implementations of
frequency for TRACE.]
- Spivey, M., Grosjean, M., & Knoblich, G. (2005). Continuous
attraction toward phonological competitors. Proceedings of the
National Academy of Sciences, 102(29), 10393-10398. [Links
TRACE to a recurrence-normalization model to predict continuuous
mouse movements linked to on-line spoken word
recognition.]
Known bugs and workarounds
If you encounter a bug in jTRACE, please carefully document the problem and
the circumstances under which it occurs. Please email bug reports to
Jim Magnuson, with 'jtrace bug' in
the subject line. Known bugs will be resolved in order of priority.
- Help window doesn't appear.
- If the help documentation does not open when
told, navigate to the jTRACE/doc folder and open manual.html file in a web browser.
- Scripting
- Some edits made during scripting do not cause an update of the interface.
In other words, after you change something, you don't see the result of it. This
problem usually affects adding and deletion of items. As a workaround, to see the
results of these changes, it is necessary to save and reload the script file.
- I run the script, but nothing happens. Currently, jTRACE does not validate
the contents of a script when it is run. Save and reload the script, double-check the
logic of the script. If it still doesn't work, let us know.
3rd Party Software
jTRACE is coded in Java. The Java Web Services Developers Pack
(JWSDP 1.5) is used. Two third party libraries are also included.
Licenses governing their use are included in the
THIRDPARTYLICENSES.txt file, included with the jTRACE download.
Permission to use these tools has greatly enhanced jTRACE
functionality.
License
This software is copyright under the following conditions:
Permission to use, copy, and modify this software and its
documentation for any purpose other than distribution-for-profit is
hereby granted without fee, provided that the above copyright notice
and this permission notice appear in all copies of the software and
related documentation.
Permission to distribute the software or modified or extended versions
thereof on a not-for-profit basis is explicitly granted, under the
above conditions. HOWEVER, THE RIGHT TO DISTRIBUTE THE SOFTWARE OR
MODIFIED OR EXTENDED VERSIONS THEREOF FOR PROFIT IS *NOT* GRANTED
EXCEPT BY PRIOR ARRANGEMENT AND WRITTEN CONSENT OF THE COPYRIGHT
HOLDERS.
THE SOFTWARE IS PROVIDED "AS-IS" AND WITHOUT WARRANTY OF ANY KIND,
EXPRESS, IMPLIED OR OTHERWISE, INCLUDING WITHOUT LIMITATION, ANY
WARRANTY OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
IN NO EVENT SHALL THE UNIVERSITY OF CONNECTICUT BE LIABLE FOR ANY
SPECIAL, INCIDENTAL, INDIRECT OR CONSEQUENTIAL DAMAGES OF ANY KIND, OR
ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS,
WHETHER OR NOT ADVISED OF THE POSSIBILITY OF DAMAGE, AND ON ANY THEORY
OF LIABILITY, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
PERFORMANCE OF THIS SOFTWARE.