It is well recognized that the physiological/pathological consequences of an immune response, against a foreign or a self‐antigen, are often critically dependent on the class of immunity generated. Here we focus on how antigen interacts with the cells of the immune system to determine whether antigen predominantly generates Th1 or Th2 cells. We refer to this mechanism as the ‘decision criterion’ controlling the Th1/Th2 phenotype of the immune response. A plausible decision criterion should account for the variables of immunization known to affect the Th1/Th2 phenotype of the ensuing immune response. Documented variables include the nature of the antigen, in terms of its degree of foreignness, the dose of antigen and the time after immunization at which the Th1/Th2 phenotype of the immune response is assessed. These are quantitative variables made at the level of the system. In addition, the route of immunization is also critical. I describe a quantitative hypothesis as to the nature of the decision criterion, referred to as the Threshold Hypothesis. This hypothesis accounts for the quantitative variables of immunization known to affect the Th1/Th2 phenotype of the immune response generated. I suggest and illustrate how this is not true of competing, contemporary hypotheses. I outline studies testing predictions of the hypothesis and illustrate its potential utility in designing strategies to prevent or treat medical situations where a predominant Th1 response is required to contain an infection, such as those caused by HIV‐1 and by Mycobacterium tuberculosis, or to contain cancers.

How an antigen interacts differently with lymphocytes and other cells of the immune system, to result in the generation of distinct classes of immunity, is one of the most basic questions of immune regulation. Understanding the nature of these “decision criteria” is central to developing effective medical interventions. Clinical observations lead to the recognition that much disease is due to an inappropriate class of immunity being generated, inappropriate because damaging, as in autoimmunity and allergies, or inappropriate because ineffective, against pathogens and cancer. I argue that the prevalent, contemporary conceptual frameworks, employed to analyze the nature of the decision criterion controlling the Th1/Th2 phenotype of the immune response, are implausible, as they ignore pertinent, classical observations. I outline reasons for favoring a different framework, that takes these observations into account, and explore its pertinence to the design of strategies of medical intervention.

Recent reports suggest a quorum of T cells is required to activate T lymphocytes and that this requirement may help explain why scarce lymphocytes, specific for peripheral self-antigen, are rarely activated by Ag. This proposal runs counter to the commonly held framework that the Ag-dependent, but CD4 T lymphocyte–independent, activation of CD8 T lymphocytes, and the activation of CD4 T lymphocytes themselves, can occur when a single CD8 or CD4 T lymphocyte encounters Ag under appropriately dangerous circumstances. We argue that a review of older literature often ignored, as well as of contemporary studies, supports the quorum concept and is difficult to reconcile with the Danger Model.

The first ideas leading to The Two‐Signal Model of lymphocyte activation were published 50 years ago, but the model was not realized in one sitting. I describe the three phases that led to its contemporary formulations. A motivation underlying all these models was to generate a minimal description of what is required for antigen to inactivate and activate mature lymphocytes that, at the same time, accounts for how peripheral tolerance is achieved. I suggest the two signal model has not only provided a substantiated framework for understanding how antigen interacts differently with B cells and CD8 T cells, to result in their inactivation and activation, but its postulates are pertinent to contemporary issues concerning the inactivation and activation of CD4 T cells.

The establishment of central tolerance to most self‐antigens results in a repertoire of mature peripheral lymphocytes specific for foreign and peripheral self‐antigens. The framework that single, mature lymphocytes are inactivated by antigen, whereas their activation requires lymphocyte cooperation, accounts for diverse observations and incorporates a mechanism of peripheral tolerance. This framework accounts for the generalizations that the sustained activation by antigen of most B cells and CD8 T cells requires CD4 T helper cells; in the absence of CD4 T cells, antigen can inactivate these B cells and CD8 T cells. In this sense, CD4 T cells are the guardians of the fate of most B cells and CD8 T cells when they encounter antigen. I argue here that the single‐lymphocyte/multiple‐lymphocyte framework for the inactivation/activation of lymphocytes also applies to CD4 T cells. I consider within this framework a model for the activation/inactivation of CD4 T cells that is consistent with the large majority of contemporary observations, including significant clinical observations. I outline the grounds why I feel this model is more plausible than the contemporary and predominant pathogen‐associated molecular pattern (PAMP) and Danger Models for CD4 T cell activation. These models are based upon what I consider the radical premise that self–nonself discrimination does not exist at the level of mature CD4 T cells. I explain why I feel this feature renders the PAMP and Danger Models somewhat implausible. The model I propose, in contrast, is conservative in that it embodies such a process of self–nonself discrimination.