In his most recent work, Sangiovanni has retreated from his stronger claims about practice-dependence. Instead of claiming that principles of justice must be practice-dependent, he now expresses his claim in a modal form, arguing that there are several ways in which practices may matter. While merely mapping out the logical space of possibilities seems to look like a modest ambition, the conditions for when practices do matter according to Sangiovanni's analysis are easily met in actuality. Consequently, if he is right, the practice-dependent approach covers a significant number of political theories. Sangiovanni's main claim is that higher-level principles with an open texture, which include most higher-level principles in political philosophy, justify a practice-dependent method in the form of a mode of application called 'mediated deduction,' according to which a thoroughgoing investigation is made of the nature of the target practice. Our task in this paper is to reject this claim. This is done in two steps. First, we question Sangiovanni's distinction between instrumental application and mediated deduction, arguing that it remains unclear whether it marks out two sufficiently distinct 'modes' to do any theoretical work. Second, we argue that the practice-dependent method is not required even if two such modes are established.

A common denominator of recent proposals suggested by political realists has been a rather pessimistic view of what we may rightfully demand of political authorities in terms of legitimacy. In our analysis, three main justificatory strategies are utilized by realists, each supposedly generating normative premises for this low bar conclusion. These strategies make use of the concept of politics, the constitutive features of politics, and feasibility constraints, respectively. In this article, we make three claims: first, that the two justificatory strategies of utilizing the concept of politics and the constitutive features of politics fail, since they rely on implausible normative premises; second, that while the feasibility strategy relies on reasonable premises, the low bar conclusion does not follow from them; third, that relativist premises fit better with the low bar conclusion, but that this also makes the realist position less attractive and casts doubt on several of its basic assumptions.

This article proposes a new definition of information security, the ‘Appropriate Access’ definition. Apart from providing the basic criteria for a definition—correct demarcation and meaning concerning the state of security—it also aims at being a definition suitable for any information security perspective. As such, it bridges the conceptual divide between so-called ‘soft issues’ of information security (those including, e.g., humans, organizations, culture, ethics, policies, and law) and more technical issues. Because of this it is also suitable for various analytical purposes, such as analysing possible security breaches, or for studying conflicting attitudes on security in an organization. The need for a new definition is demonstrated by pointing to a number of problems for the standard definition type of information security—the so-called CIA definition. Besides being too broad as well as too narrow, it cannot properly handle the soft issues of information security, nor recognize the contextual and normative nature of security.

A shared presumption among practice-dependent theorists is that a principle of justice is dependent on the function or aim of the practice to which it is supposed to be applied. In recent contributions to this debate, the condition of epistemic uncertainty plays a significant role for motivating and justifying a practice-dependent view. This paper analyses the role of epistemic uncertainty in justifying a practice-dependent approach. We see two kinds of epistemic uncertainty allegedly playing this justificatory role. What we call ‘normative epistemic uncertainty’ emerges from dealing with the problem of value uncertainty in justifying applied principles when our higher-level principles are open-textured, that is, when their content is too vague or unclear to generate determinate prescriptions. What we call ‘descriptive epistemic uncertainty’ emerges from dealing with uncertainty about empirical facts, such as the problem of moral assurance, that is, the problem that the requirements of justice cannot go beyond arrangements that we can know with reasonable confidence that we can jointly establish and maintain. In both cases, practice-dependent theorists conclude that the condition of epistemic uncertainty justifies a practice-dependent approach, which puts certain restrictions on theorizing regulative principles and has wide-ranging practical implications for the scope of justice. Our claim in this paper is that neither kind of epistemic uncertainty justifies a practice-dependent approach.

Many political theorists in current debates have argued that pragmatist theories of mind and language place certain constraints on our normative political theories. In a couple of papers, we have accused these pragmatically influenced political theorists of misapplication of otherwise perfectly valid ideas. In a recent paper, one of the targets of our critique, Thomas Fossen, has retorted that we have misrepresented the role that a pragmatist theory of language plays in these accounts. In this paper, we claim that Fossen’s attempt to chisel out a role for his account in normative political theory rehearses the same problematic view of the utility of theories of language as his previous iterations. We argue that Fossen’s account is still guilty of the fallacious claim that a pragmatist theory of language (in his case Robert Brandom’s account) has implications for the form and justification of theories of political legitimacy. We specifically focus on three flaws with his current reply: the idea that criteria and conditions are problematic on a pragmatist outlook, the idea that a pragmatist linguistic account applied to a particular political context will have a distinct political-theoretical payoff, and the idea that a fundamental linguistic level of analysis supplies normative guidance for theorizing political legitimacy.

The concepts of risk, safety, and security have received substantial academic interest. Several assumptions exist about their nature and relation. Besides academic use, the words risk, safety, and security are frequent in ordinary language, for example, in media reporting. In this article, we analyze the concepts of risk, safety, and security, and their relation, based on empirical observation of their actual everyday use. The "behavioral profiles" of the nouns risk, safety, and security and the adjectives risky, safe, and secure are coded and compared regarding lexical and grammatical contexts. The main findings are: (1) the three nouns risk, safety, and security, and the two adjectives safe and secure, have widespread use in different senses, which will make any attempt to define them in a single unified manner extremely difficult; (2) the relationship between the central risk terms is complex and only partially confirms the distinctions commonly made between the terms in specialized terminology; (3) whereas most attempts to define risk in specialized terminology have taken the term to have a quantitative meaning, nonquantitative meanings dominate in everyday language, and numerical meanings are rare; and (4) the three adjectives safe, secure, and risky are frequently used in comparative form. This speaks against interpretations that would take them as absolute, all-or-nothing concepts.

In many decision-situations, we are uncertain not only about the facts but also about our own values that we intend to apply to the problem. Which values are at stake, and whether and how those values compare may not always be clear to us. This chapter introduces the issue and discusses some ways to deal with value uncertainty in practical decision-making. In particular, four types of uncertainty of values are introduced: uncertainty about which values we endorse, uncertainty about the specific content of the values we do endorse, uncertainty about which among our values apply to the problem at hand, and the relative weight among different values we endorse. Various ways of contributing to solving value uncertainty are then discussed: contextualization, hierarchy of values, assigning strength to values, embedding and transforming the problem. Furthermore, two methods of dealing with value uncertainty remaining even after these methods have been applied are treated.