Theory of Multiple Intelligences: Is It a Scientific Theory?

by Jie-Qi Chen - 2004

This essay discusses the status of multiple intelligences (MI) theory as a scientific theory by addressing three issues: the empirical evidence Gardner used to establish MI theory, the methodology he employed to validate MI theory, and the purpose or function of MI theory.


Of critical importance to the scientific establishment of a theory is the methodology by which the theory is created and developed (Kuhn, 1962). This principle became clearer as the fields of the history and philosophy of science matured. Before the mid-20th century, philosophers of science such as Karl Popper (1959) attempted to define an objective and universally applicable methodology for all sciences. Such an attempt inevitably failed because it did not recognize the necessary interconnection between methodologies and the objects of study. By the 1960s, scholars in the history and philosophy of science agreed that the methodologies of science both shape and are shaped by the subjects they are applied to. Thus, the absolute objectivity of any methodology is illusionary (Kuhn, 1962). The history of this debate about methodology parallels the debate regarding the scientific credibility of MI theory.

Since the inception of MI theory, some scholars in the field of cognitive psychology have questioned its status as a scientific theory. Specific criticisms of the theory include the following: ‘‘This looks like pop psychology’’; ‘‘There has been no empirical data to validate the theory’’; and ‘‘The independence of multiple intelligences has not been tested empirically’’. In the process of developing MI theory, Gardner (1993a) considered the range of adult end-states that are valued in diverse cultures around the world. To identify the abilities that support these end-states, he examined empirical data from disciplines that had not been considered previously for the purpose of defining human intelligence. The results of Gardner’s analyses consistently supported his emerging notion of specific and relatively independent sets of cognitive abilities. His examination of these data sets also yielded eight criteria for identifying an intelligence. To be defined as an intelligence, an ability has to be tested in terms of the following eight criteria:

An intelligence should be isolable in cases of brain damage and there should be evidence for its plausibility and autonomy in evolutionary history. These two criteria were derived from biology.

Two criteria came from developmental psychology: An intelligence has to have a distinct developmental history with a definable set of expert end-state performances and it must exist within special populations such as idiot savants and prodigies.

Two criteria emerged from traditional psychology: An intelligence needs to demonstrate relatively independent operation through the results of specific skill training and also through low correlation to other intelligences in psychometric studies.

Two criteria were derived from logical analysis: An intelligence must have its own identifiable core operation or set of operations and must be susceptible to encoding in a symbol system such as language, numbers, graphics, or musical notations.

Although Gardner (1993a) did not base his theory on testing of children and statistical analyses of the results, the primary method used by psychometricians to establish the credibility of the construct of IQ, he did ground the theory on analysis of empirical data. The eight criteria used to identify intelligences are not the reverie of a giant mind. Rather they are derived from Gardner’s comprehensive, thorough, and systematic review of empirical data from studies in biology, neuropsychology, developmental psychology, and cultural anthropology.

Because MI theory is based on the conception of human cognitive functioning in diverse real-life situations, its scientific establishment is grounded in empirical data that describe the functioning of multiple abilities in diverse situations. For example, MI theory better accounts for data that describe the cognitive functions of special population than intelligence defined as IQ does. Exhibiting differentiated profiles of specific abilities, these populations include those who have suffered brain injury as well as prodigies and savants. MI theory also better describes various learning profiles that teachers and educators encounter on a daily basis. Finally, MI theory better explains the diverse abilities required to succeed in different professions.

Clearly, the scientific evidence used to support the psychometric notion of intelligence as IQ and the evidence used to establish MI theory are radically different. Referring back to lessons learned from the history of science, there can and should be more than one way to study human intelligence. If we limit studies by relying on a single standard for the acceptable measurement of intelligence, our understanding of this most central capacity of human beings will be significantly restrained.


In the field of psychology, a theory of intelligence is typically validated by establishing two psychometric properties of tests based on the theory: validity and reliability. Validity refers to the degree to which a test measures its intended attributes or desired outcomes. Although there are many kinds of validity, the most commonly reported in the manual of standardized intelligence tests is concurrent validity. It is usually established by comparing scores on one test with scores of other standardized tests of the same nature. Reliability refers to the consistency of a test’s result over time and is usually determined by using one or more of the following methods: test-retest, equivalent-form, and split-half. Correlation is the statistical technique that almost all standardized intelligence tests use to report the degree of validity and reliability.

Validity and reliability are useful measures for testing the theoretical construct that human intelligence is a general ability that remains stable over time. If human intelligence is a general ability, different measures of the ability should be positively correlated (validity). Since the general ability is stable, measures of it at different times should be correlated as well (reliability). Defining and measuring intelligence using IQ tests makes it possible to rank order individuals based on a single numerical score that is expected to remain constant. Scores on IQ tests can also be used to categorize individuals based on the amount of intelligence they possess.

The means used to validate a theory are shaped by the constructs and uses of the theory. Gardner (1993a) argues that human intelligence is not a general ability. Rather, it is a biopsychological potential with an emergent, responsive, and pluralistic nature. To validate this theoretical construct, one has to develop means radically different from intelligence tests. As Vygotsky (1978) argued, ‘‘Any fundamentally new approach to a scientific problem inevitably leads to new methods of investigation and analysis. The invention of new methods that are adequate to the new ways in which problems are posed requires far more than a simple modification of previously accepted methods’’ (p. 58).

 If we were developing a psychological assessment to test multiple intelligences what would its critical features be? For one, accurate assessment of multiple intelligences demands a range of measures that tap the different facets of each intellectual capacity. Also, intelligence-fair instruments are needed to assess the unique faculties of each intelligence. Intelligence-fair instruments engage the key components of particular intelligences, allowing one to look directly at the functioning of each intellectual capacity. Further, the assessment must be an ongoing process based on multiple samples of an individual’s abilities over time in different contexts, taking into consideration the child’s educational and cultural experiences. Finally, assessments of multiple intelligences are designed to identify and build on individuals’ strengths by creating rich educational environments with learning opportunities that match children’s specific abilities and interests (Chen & Gardner, 1997). Needless to say, the development of such assessments requires concerted efforts over a long period of time to produce quality instruments and to carefully train individuals who can administer and interpret them in a sensitive manner (Adams, 1993; Hsueh, 2003; Krechevsky, 1998; McNamee, Chen, Masur, McCray, & Melendez, 2003; Shearer, 1996; Yoong, 2001).

The methods used to validate MI theory are not limited to the development of new psychological assessments. MI theory can also be validated by evaluating the results of applying the theory in a range of educational settings. Many articles in this volume indicate that both teachers and parents have consistently reported that MI theory has given them more accurate perceptions of children’s intellectual potentials and more specific methods for supporting and developing these potentials. Kornhaber and her colleagues at Harvard University’s Project Zero studied 41 elementary schools in the United States that had applied MI theory to school-based practices for at least 3 years. Among schools that reported improvements in standardized-test scores, student discipline, parent participation, or the performance of students with learning differences, the majority linked the improvements to MI-based interventions (Kornhaber, 1999; Kornhaber, Veenema, & Fierros, 2003). The effectiveness of these applications is an important source for the validation of MI theory.


Whether a theory has value to a specific field or society depends on the explanatory power and the generative power of the theory (Kuhn, 1962; Losee, 1980). A theory that has high explanatory power can account for a wide range of observations. It brings order and coherence to information, clarifying the relations of parts to whole, and describing underlying mechanisms. A theory that has high generative power orients investigators to the future by offering new frameworks for studying unknowns and contributing new knowledge to the field. It stimulates new ideas, provides new questions, and leads to new ways of understanding the world.

Both the explanatory and the generative power of MI theory are high. As described earlier, the amount and range of empirical evidence that Gardner (1993a) cites and synthesizes in making his case for MI theory is substantial. Further, MI theory makes sense to practitioners and fits their experience about individuals’ intellectual strengths and weaknesses. That it makes sense is clear evidence of the explanatory power of MI theory.

Although not all psychologists agree with Gardner’s theory of eight relatively independent intelligences, Gardner’s claim that the nature of the human intelligences is emergent, responsive, and pluralistic is no longer a novel idea in the field of cognitive psychology. MI theory has contributed to changing our perception and understanding of human intelligences. Due to its high generative power, MI theory has stimulated countless new ideas and practices in the field of education (Campbell, Campbell, & Dickinson, 1996; Chen, 1993; Chen, Krechevsky, & Viens, 1998; Kornhaber, 1999; Kornhaber, Veenema, & Fierros, 2003; Lazear, 1994; New City School, 1994).

In the area of curriculum development, for example, MI-based curricula encompass a broad range of subject areas that include but go beyond skill development in reading, writing, and arithmetic. Because all intelligences are equally valuable, subjects such as visual arts and creative movement are also included in the curriculum. According to MI theory, the talented artist and the developing dancer are just as intelligent as the excellent reader, and each has an important place in society. Also, an authentic MI-based approach goes beyond learning factual knowledge. It also stresses the importance of promoting in-depth exploration and real understanding of the key concepts essential to a domain (Gardner, 2000).

Educators who work with at-risk children have been particularly drawn to the application of MI theory because it offers an approach to intervention that focuses on strengths instead of deficits. By the same token, MI theory extends the concept of the gifted child beyond those who excel in linguistic and logical pursuits to include children who achieve in a wide range of domains.

MI theory can be applied to the development of instructional techniques as well. For example, a teacher can provide multiple entry points to the study of a particular topic by using different media and encouraging students to express their understanding of the topic through diverse representational methods such as writing, three-dimensional models, or dramatizations. Such instructional approaches make it possible for students to find ways of learning that are attuned to their predispositions and therefore increase their motivation and engagement in the learning process. Use of these approaches also increases the likelihood that every student will attain some understanding of the topic at hand.

In summation, in discussions of whether MI theory is a scientific theory, two points warrant special attention. First, intelligence is not a tangible object that can be measured; it is a construct that psychologists define. As theoretical hypotheses differ, so does the methodology used to develop the theory and the evidence cited to validate the theory. Any attempt to apply a uniform standard for establishing the credibility and value of a theory at best fails to consider the possibility of alternative approaches, and at worst impedes the development of new ideas by constraining the use of new methodologies. Second, theories, particularly theories in the social sciences, are rarely proved or disapproved decisively, regardless of the methodology used to test the theoretical construct. A theory is not necessarily valuable because it is supported by the results of experimental tests. Rather, its value depends on the contributions it makes to understanding and to practice in the field. The value of MI theory has been clearly established by its many successful applications in the field.


Adams, M. (1993). An empirical investigation of domain-specific theories of preschool children’s cognitive abilities. Unpublished doctoral dissertation, Tufts University, Medford, MA.

Campbell, L., Campbell, B., & Dickinson, D. (1996). Teaching and learning through multiple intelligences. Needham Heights, MA: Allyn & Bacon.

Chen, J. Q. (1993, April). Working with at-risk children through the identification and nurturance of their strengths. Paper presented at the biennial conference of the Society for Research of Child Development, New Orleans, LA.

Chen, J. Q., & Gardner, H. (1997). Alternative assessment from a multiple intelligences theoretical perspective. In D. P. Flanagan, J. L. Genshaft, & P. L. Harrison (Eds.), Beyond traditional intellectual assessment: Contemporary and emerging theories, tests, and issues (pp. 105–121). New York: Guilford.

Chen, J. Q., Krechevsky, M., & Viens, J. (1998). Building on children’s strengths: The experience of Project Spectrum. New York: Teachers College Press.

Gardner, H. (1993a). Frames of mind: The theory of multiple intelligences (10th-anniversary ed.). New York: Basic Books.

Gardner, H. (1993b). Multiple intelligences: The theory in practice. New York: Basic Books.

Gardner, H. (1999). Intelligence reframed: Multiple intelligences for the 21st century. New York: Basic Books.

Gardner, H. (2000). The disciplined mind: Beyond facts and standardized tests, the K–12 education that every child deserves. New York: Penguin Books.

Hsueh, W. C. (2003, April). The development of a MI assessment for young children in Taiwan. Paper presented at the annual meeting of the American Educational Research Association, Chicago, IL.


Kornhaber, M. (1999). Multiple intelligences theory in practice (pp. 179–191). In J. Block, S. T. Everson, & T. R. Guskey (Eds.), Comprehensive school reform: A program perspective. Dubuque, IA: Kendall/Hunt Publishers.

Kornhaber, M., Veenema, S., & Fierros, E. (2003). Multiple intelligences: Best ideas from research and practice. Boston: Allyn and Bacon.

Krechevsky, M. (1998). Project Spectrum preschool assessment handbook. New York: Teachers College Press.

Kuhn, T. (1962). The structure of scientific revolution. Chicago: University of Chicago Press.

Lazear, D. (1994). Seven pathways of learning: Teaching students and parents about multiple intelligences. Tucson, AZ: Zephyr.

Losee, J. (1980). A historical introduction to the philosophy of science (2nd ed.). Oxford: University Press.

McNamee, G., Chen, J. Q., Masur, A., McCray, J., & Melendez, L. (2002). Assessing and teaching diverse learners. Journal of Early Childhood Teacher Educators, 23(3), 275–282.

New City School. (1994). Multiple intelligences: Teaching for success. St. Louis, MO: Author.

Popper, K. (1959). The logic of scientific discovery. London: Hutchison.

Shearer, B. (1996). The MIDAS: professional manual. Kent, OH: MI Research and Consulting.

Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes (M. Cole, V. John-Steiner, S. Scribner, & E. Souberman, Trans.). Cambridge, MA: Harvard University Press.

Yoong, S. (2001, November). Multiple intelligences: A construct validation of the MIDAS Scale in Malaysia. Paper presented at the International Conference on Measurement and Evaluation in Education, Penang, Malaysia.

Cite This Article as: Teachers College Record Volume 106 Number 1, 2004, p. 17-23 ID Number: 11505, Date Accessed: 1/23/2022 8:10:21 PM

Purchase Reprint Rights for this article or review