JID:PLREV AID:234 /DIS[m3SC+; v 1.137; Prn:23/12/2011; 11:27] P.1(1-2)
Available online at www.sciencedirect.com
Physics of Life Reviews ••• (••••) •••–•••
Nonlinear dynamics and higher cognitive mental functions
Comment on “Information flow dynamics in the brain”
by M.I. Rabinovich et al.
Harvard University, Athinoula A. Martinos Center for Biomedical Imaging, 149 Thirteenth Street, Charlestown, MA 02129, United States
Received 20 December 2011; accepted 20 December 2011
Communicated by L. Perlovsky
In the target article  M. Rabinovich and co-authors reviewed a theory of nonlinear dynamic of information flow
in the brain. They described how mental representations can be modeled in this theory. They addressed cognitive
processes of top-down and bottom-up information flow, which are fundamental for relating information inside the
brain to objects and events in the outside world. The article appropriately emphasizes that Shannon information is
inadequate for characterizing conceptual contents of mental representations.
For example, a picture containing one million pixels with values (0,1) contains 1 million bits of Shannon’s infor-
mation. However this value tells us nothing about presence or absence of food or predators, information essential for
life. Not only Shannon’s theory of information cannot measure presence or absence of life-essential information, it
also contains no algorithms for learning such measures.
The target article states that Kolmogorov and Chaitin information measures are adequate for characterizing con-
ceptual contents of mental representations. I am not convinced and would suggest that this statement requires further
elaboration and likely remains a topic for future research.
Humans as well as higher animals can perceive objects (including food and predators), in other words can represent
in their minds information content of objects. This is based partly on evolutionary evolved mental properties and partly
on learning from objects that can be perceived in the surrounding world. This learning aspect seems to be absent from
Shannon, Kolmogorov, or Chaitin theories.
Humans possess unique ability to learn abstract concepts-representations. This human ability has not been ex-
plained in previous cognitive research. A recent hypothesis attempting such an explanation  suggests that ability
for abstract thinking is inextricably connected to language. The mental hierarchy is modeled in  as a dual hierarchy
of language and cognition representations. Language representations are learned by human children at an early age
throughout the entire hierarchy. This is possible because language representations exist in the surrounding language
“ready-made” . But cognitive representations for abstract concepts do not exist ready-made and require learning
from real-life experience in correspondence to language contents. I am looking forward to further development of the
nonlinear dynamic theory in the target article  toward modeling the dual hierarchy of language and cognition.
DOI of original article: 10.1016/j.plrev.2011.11.002.
E-mail address: email@example.com.
1571-0645/$ – see front matter © 2011 Published by Elsevier B.V.
JID:PLREV Download full-text
AID:234 /DIS[m3SC+; v 1.137; Prn:23/12/2011; 11:27] P.2(1-2)
L. Perlovsky / Physics of Life Reviews ••• (••••) •••–•••
 Rabinovich MI, Afraimovich VS, et al. Information flow dynamics in the brain. Physics of Life Reviews 2012 [in this issue].
 Perlovsky LI. Physics of Life Reviews 2006;3(1):23–55.
 Perlovsky LI, Ilin R. Neurally and mathematically motivated architecture for language and thought. Special issue “Brain and Language Archi-
tectures: Where We are Now?”. The Open Neuroimaging Journal 2001;4:70–80. http://www.bentham.org/open/tonij/openaccess2.htm.