The content of the paper refers to the field of interdisciplinary scientific research, lying at the intersection of information theory and cognitive psychology.

Relevance. In conditions of rapid development of communication between people through computers, the transfer and adequate reception of the meaning of an information message are of great importance.

Purpose. Theoretical substantiation and development of the probabilistic approach to the evaluation of the meaning of purposeful human activity on the example of a text.

Methods. Theoretical analysis of probabilistic (V. Nalimov) and activity-related (D. Leontiev) approaches to the problem of meaning in psychology; method of calculating the amount of information developed by K. Shannon; author’s technique for assessing the meaning of a text message, based on a comparative analysis of the source text with its probabilistic representation by random permutation of words.

Results.The justification of the probabilistic nature of sense is given; a psychometric criterion is proposed for quantifying the meaning of purposeful conscious human activity in the form of differential entropy of the exponential distribution of the Shannon entropy difference between two adjacent words in the text; a technique for probabilistic evaluation of the meaning of purposeful conscious human activity taking text as an example is proposed.

The analysis of more than 50 texts - information messages from the Internet, differing in the level of meaningfulness and reflecting various emotional experiences of the authors - showed: a) the greatest meaningfulness is in texts written by authors under the influence of strong emotions; b) texts the style of presentation of which is typical for authors in the state of neutral emotions occupy an intermediate position in terms of meaningfulness; c) the least meaning is in texts written by those who experience some disorder of thinking.

Conclusions. The results confirm the main hypothesis of the study that the meaning is encoded in a chain of semantic connections of purposeful conscious human activity and has a probabilistic nature. The materials of the paper can be of practical and theoretical interest for psychologists dealing with issues of meaning, linguists and developers of artificial intelligence.

The discussion of crucial problems of psychophysiology and natural sciences is presented, in general: are there any recurrences of measured values of psycho­physiological parameters (functions) of a person. It is shown that it is impossible to obtain arbitrarily a repetition of samples for tremor and electromyogram (when trying to maintain static muscular tension), statistical distribution function of samples f(x) is continuously changing. This dynamics is typical for tremor (conditionally involuntary movement) and electromyogram when the higher nervous activity of individual oversees the muscular effort. The involuntary movement (tremor) shows the number of coincidence of electromyogram samples (at unrestricted muscular tension) k, which is similar to such in the matrices of pairwise comparisons of tremor samples. Shannon’s entropy in this case does not change significantly.