Abstract
The article proceeds upon the assumption that the beliefs and degrees of belief of rational agents satisfy a number of constraints, including: (1) consistency and deductive closure for belief sets, (2) conformity to the axioms of probability for degrees of belief, and (3) the Lockean Thesis concerning the relationship between belief and degree of belief. Assuming that the beliefs and degrees of belief of both individuals and collectives satisfy the preceding three constraints, I discuss what further constraints may be imposed on the aggregation of beliefs and degrees of belief. Some possibility and impossibility results are presented. The possibility results suggest that the three proposed rationality constraints are compatible with reasonable aggregation procedures for belief and degree of belief.
Similar content being viewed by others
Notes
For an excellent survey of the topic, see Pigozzi (2015).
It may be that we could integrate individual beliefs in order to form collective degrees of belief. I will not explore that possibility here.
The project of the present paper is similar to a project that was briefly outlined by Cariani (2016, pp. 402––403). Cariani considers two alternate approaches to belief aggregation, namely, (1) a Bayesian model that updates a prior probability function by conditionalization on expert testimony in order to form a posterior probability function to which an acceptance rule is then applied, and (2) a more standard aggregation model applied to the testimony/beliefs of the experts. Cariani considers the issue of whether the two approaches would agree in their results, which parallels the problem that is the focus of the present paper. The project envisioned by Cariani differs from the one pursued here, since (among other differences) the Bayesian approach considered by Cariani differs markedly from the approach to probability aggregation considered here.
I assume that the domain of Fbel includes ordered sets of belief functions of different size (n ≥ 1) over different size partitions of W (k ≥ 1).
I here follow the definition of Pauly and Van Hees (2006, p. 573).
A proof of this fact, along with other theorems, is given in “Appendix B”.
I assume that the domain of Fprob includes ordered sets of probability functions of different size (n ≥ 1) over different size partitions of W (k ≥ 1).
But see Genest and Wagner (1984), who introduce grounds for doubting Zp.
For a proof, see Lehrer and Wagner (1983).
As observed by Russell, Hawthorne, and Buchak (2015), violations of CL can also lead a group to accept a diachronic Dutch book, assuming (among other things) that the group re-aggregates their degrees of belief each time they are faced with making a collective decision. I do not believe that the observation of Russell, Hawthorne, and Buchak provides a decisive reason in favor of CL. In the kind of example presented by Russell, Hawthorne, and Buchak, a group accepts a diachronic Dutch book, in a case where each agent updates her degrees of belief by conditionalizing on the same proposition. This possibility would be blocked, if the group’s aggregation procedure had satisfied CL. However, cases where each agent in a group updates her degrees of belief by conditionalizing on the same proposition are not typical, and it is clear that conformity to CL will not prevent a group from accepting a diachronic Dutch book in cases where members of the group update on different propositions. I doubt that there is any plausible aggregation principle that will protect a group from accepting a diachronic Dutch book, in situations where the group members update on different propositions, suggesting that we must look beyond aggregation principles, such as CL, as a means of preventing collectives from accepting diachronic Dutch books.
As a point of comparison, “Appendix A” considers Lin and Kelly’s (2012) symmetric camera shutter rule. The contents of “Appendix A” show that, for the most part, the symmetric camera shutter rule does not offer the opportunity of escaping impossibility results that arise for the Lockean Thesis.
If we consider the least restriction of the domain of Fprob that is consistent with geometric and multiplicative opinion pooling (requiring at least one j, such that, for all i, rij > 0), then these two sorts of pooling are also inconsistent with the conjunction of Ub, Ab, Nb, and PLC. The preceding is demonstrable by the example of B = 〈〈0, 0, 1〉, 〈0, 1, 0〉, 〈1, 0, 0〉〉 and P = 〈〈0, 0.1, 0.9〉, 〈0.1, 0.9, 0〉, 〈0.9, 0, 0.1〉〉. It is an open question whether some natural restriction on the domain of Fprob (e.g., for all i and j, rij > 0) would be sufficient to make some form of geometric or multiplicative opinion pooling consistent with Ub, Ab, Nb, and PLC. See Dietrich (2010) and Dietrich and List (2016) for excellent discussions of geometric and multiplicative opinion pooling.
The more general unanimity condition that applies for all values in [0, 1] (rather than merely in {0, 1}) does not cohere with PLC, in the presence of other reasonable principles.
For all i: simply let pP(wi) = 1/|Π| and bB(wi) = 1.
For the proof that {Ub, Ab, Nb, Zb, Up, SFr, CL} is inconsistent, for given r in (0, 1), simply select the least z such (1 − z)/z ≤ 1 − r, and then the least y such that y + x = 1 – z and x/y ≤ 1 − r, and let P = 〈〈x, y, z〉, 〈z, y, x〉〉.
For the proof that {Ub, Ab, Nb, Zb, Up, Ap, Np, SFr, IP} is inconsistent, for given r in (0, 1), simply select the least z such (1 − z)/z ≤ 1 − r, and let P = 〈〈x, √x − x, √x − x, z〉, 〈z, √x − x, √x − x, x〉〉.
References
Arló-Costa, H., & Pedersen, A. (2012). Belief and probability: A general theory of probability cores. International Journal of Approximate Reasoning, 53, 293–315.
Briggs, R., Cariani, F., Easwaran, K., & Fitelson, B. (2014). Individual coherence and group coherence. In J. Lackey (Ed.), Essays in collective epistemology (pp. 215–249). Oxford: Oxford University Press.
Brössel, P., & Eder, A. (2014). How to resolve doxastic disagreement. Synthese, 191, 2359–2381.
Buchak, L. (2014). Belief, credence, and norms. Philosophical Studies, 169, 1–27.
Cariani, F. (2016). Local supermajorities. Erkenntnis, 81, 391–406.
Chandler, J. (2013). Acceptance, aggregation and scoring rules. Erkenntnis, 78, 201–217.
Dietrich, F. (2010). Bayesian group belief. Social Choice and Welfare, 35, 595–626.
Dietrich, F., & List, C. (2016). Probabilistic opinion pooling. In C. Hitchcock & A. Hajek (Eds.), Oxford handbook of probability and philosophy. Oxford: Oxford University Press.
Dietrich, F., & List, C. (2018). From degrees of belief to binary beliefs: Lessons from judgment-aggregation theory. Journal of Philosophy, 115, 225–270.
Douven, I., & Romeijn, J. (2007). The discursive dilemma as a lottery paradox. Economics and Philosophy, 23, 301–319.
Douven, I., & Williamson, T. (2006). Generalizing the lottery paradox. British Journal for the Philosophy of Science, 57, 755–779.
Elkin, L., & Wheeler, G. (2018). Resolving peer disagreements through imprecise probabilities. Noûs, 52, 260–278.
Fantl, J., & McGrath, M. (2002). Evidence, pragmatics, and justification. Philosophical Review, 111, 67–94.
Foley, R. (1993). Working without a net. Oxford: Oxford University Press.
Foley, R. (2009). Beliefs, degrees of belief, and the lockean thesis. In F. Huber & C. Schmidt-Petri (Eds.), Degrees of belief (pp. 37–47). Dordrecht: Springer.
Ganson, D. (2008). Evidentialism and pragmatic constraints on outright belief. Philosophical Studies, 139, 441–458.
Genest, C. (1984a). A conflict between two axioms for combining subjective distributions. Journal of the Royal Statistical Society: Series B, 46, 403–405.
Genest, C. (1984b). A characterization theorem for externally Bayesian groups. The Annals of Statistics, 12, 1100–1105.
Genest, C., & Wagner, C. (1984). Further evidence against independence preservation in expert judgement synthesis. Technical Report No. 84-10. Department of Statistics and Actuarial Science, University of Waterloo.
Genest, C., & Wagner, C. (1987). Further evidence against independence preservation in expert judgement synthesis. Aequationes Mathematicae, 32, 74–86.
Genest, C., & Zidek, J. (1986). Combining probability distributions: A critique and annotated bibliography. Statistical Science, 1, 114–135.
Hawthorne, J. (2004). Knowledge and lotteries. Oxford: Oxford University Press.
Hawthorne, J. (2009). The Lockean thesis and the logic of belief. In F. Huber & C. Schmidt-Petri (Eds.), Degrees of belief (pp. 49–74). Dordrecht: Springer.
Kyburg, H. (1961). Probability and the logic of rational belief. Middletown, CT: Wesleyan University Press.
Laddaga, R. (1977). Lehrer and the consensus proposal. Synthese, 36, 473–477.
Lehrer, K., & Wagner, C. (1981). Rational consensus in science and society. Dordrecht: Reidel.
Lehrer, K., & Wagner, C. (1983). Probability amalgamation and the independence issue: A reply to Laddaga. Synthese, 55, 339–346.
Leitgeb, H. (2013). Reducing belief simpliciter to degrees of belief. Annals of Pure and Applied Logic, 164, 1338–1389.
Leitgeb, H. (2014). The stability theory of belief. Philosophical Review, 123, 131–171.
Levi, I. (2004). List and Pettit. Synthese, 140, 237–242.
Lin, H., & Kelly, K. (2012). A geo-logical solution to the lottery paradox. Synthese, 186, 531–575.
List, C., & Pettit, P. (2002). Aggregating sets of judgements: an impossibility result. Economics and Philosophy, 18, 89–110.
Pauly, M., & Van Hees, M. (2006). Logical constraints on judgement aggregation. Journal of Philosophical Logic, 35, 569–585.
Pigozzi, G. (2015). Belief merging and judgment aggregation. In E. N. Zalta (Ed.), The stanford encyclopedia of philosophy, Fall 2015 edn. http://plato.stanford.edu/archives/fall2015/entries/belief-merging/. Accessed 1 Mar 2017.
Russell, J., Hawthorne, J., & Buchak, L. (2015). Groupthink. Philosophical Studies, 172, 1287–1309.
Schurz, G. (2017). Impossibility results for rational belief. Noûs. https://doi.org/10.1111/nous.12214.
Staffel, J. (2016). Beliefs, buses and lotteries: Why rational belief can’t be stably high credence. Philosophical Studies, 173, 1721–1734.
Thorn, P. (2017). Against deductive closure. Theoria, 83, 103–119.
Wagner, C. (2010). Jeffrey conditioning and external Bayesianity. Logic Journal of the IGPL, 18, 336–345.
Weatherson, B. (2005). Can we do without pragmatic encroachment? Philosophical Studies, 19, 417–443.
Acknowledgements
Work on this paper was supported by the DFG Grant SCHU1566/9-1 as part of the priority program New Frameworks of Rationality (SPP 1516), and by the DFG Collaborative Research Centre 991: The Structure of Representations in Language, Cognition, and Science. For valuable comments and suggestions I am thankful to Peter Brössel, Ludwig Fahrbach, Theo Kuipers, Olivier Roy, Gerhard Schurz, and two anonymous reviewers for Synthese. I particularly indebted to the contributions of Christian Feldbacher-Escamilla, who came up with the idea to investigate the synchronized aggregation of beliefs and degrees of belief, and provided many valuable suggestions during conversations of this topic.
Author information
Authors and Affiliations
Corresponding author
Appendices
Appendix A: Lin and Kelly’s Symmetric Camera Shutter Rule
Motivated by the fact that there is no consistent context independent version of the Lockean Thesis with a threshold r < 1, Lin and Kelly (2012) proposed what they call the symmetric camera shutter rule. Like the Lockean Thesis the rule of Lin and Kelly includes a parameter r (0 < r < 1) that reflects an agent’s degree of credulity. Given a particular value r, the symmetric camera shutter rule demands the following relation between belief an degree of belief, where σ(wi) = p(wi)/maxj p(wj):
In line with this condition, we can say that 〈b, p〉 is shutter fit at r (abbreviated SFr(b, p)) just in case for all i: b(wi) = 0 if and only if σ(wi) ≤ 1 − r.
It is straightforward to define analogues of Lockean Credulity and PLC that apply to the symmetric camera shutter rule:
Definition
〈b, p〉 is credulous shutter fit (abbreviated CSF(b, p)) if and only if (1) there exists an r (0 < r < 1): 〈b, p〉 is shutter fit at r, and (2) for all b′: if φb′ ⊂ φb, then for all s (0 < s < 1): 〈b′, p〉 is not shutter fit at s.
Preservation of Shutter Fit Credulity (PSFC):
For all B, P: if for all i: CSF(bi, pi), then CSF(bB, pP).
It is, then, straightforward to demonstrate the analogue of Theorem 3 (and similarly for Theorems 5, 6, and 7):
Theorem 10
For all F: if {F = Fbel, Ub, Ab, Nb} is consistent, then {F = Fbel, Up, Ap, Np, PSFC} is consistent.Footnote 21
Beyond the preceding, it turns out that the symmetric camera shutter rule is ‘more stable’ than the Lockean Thesis, as demonstrated by the behavior of the following analogues of the instances of Lr:
Preservation of Shutter Fit at Level r (SFr):
For all B and P: if for all i: 〈bi, pi〉 is shutter fit at r, then 〈bB, pP〉 is shutter fit at r.
Unlike instances of Lr (for all r in (0.5, 1)), instances of SFr are consistent with {Ub, Ab, Nb, Up, Ap, Np}, and more generally:
Theorem 11
For all F, r in (0, 1): if {F = Fbel, Ub, Ab, Nb} is consistent, then {F = Fbel, Up, Ap, Np, SFr} is consistent.
However, both PSFC, along with all instances of SFr, are still incompatible with LW (in the presence of other plausible principles):
Theorem 12
{Ub, Ab, Nb, Up, Ab, Nb, LW, PSFC} is inconsistent.Footnote 22
Theorem 13
For all r in (0, 1): {Ub, Ab, Nb, Up, LW, SFr} is inconsistent.
Proof
For all r in (0, 1), we show that {Ub, Ab, Nb, Up, LW, SFr} is inconsistent. Consider an arbitrary r in (0, 1). Take the least n such that ((1 − (3√(n)/n))/(n − 1))/(3√(n)/n) ≤ 1 − r. Let |Π| = n. Let B = 〈b1, …, bn〉, where, for all i, bi has the value 1 in the ith position, and the value 0 in all other positions. Let P = 〈p1, …, pn〉, where p1 has the value v in the first position, and the value (1 − v)/(n − 1) in all other positions, where v = 3√(n)/n. For all i > 1: let pi have the value 1 in the ith position and the value 0 in all other positions. Notice that Ab and Nb imply that bB is a series of 1 s, and so bB(w1) = 1. Next notice that for all i: 〈bi, pi〉 is shutter-fit at r. (In particular, for all i > 1: σp1(wi) = (1 − (3√(n)/n))/(n − 1))/(3√(n)/n) ≤ 1 − r.) So SFr implies that 〈bB, pB〉 is shutter-fit at r. For reasons that follow, this implies that c1 > 1/n. Assume c1 ≤ 1/n, then pP(w1) ≤ (3√(n)/n)(1/n), and there exists an i such that pP(wi) ≥ (1 − (3√(n)/n))/(n − 1))(1/n) + 1/n, so that σP(w1) ≤ (3√(n)/n)(1/n)/((1 − (3√(n)/n))/(n − 1))(1/n) + 1/n). But for all n > 1: (3√(n)/n)(1/n)/((1 − (3√(n)/n))/(n − 1))(1/n) + 1/n) ≤ ((1 − (3√(n)/n))/(n − 1))/(3√(n)/n). Assume not. Then for some n > 1: 3√(n2)((n −2)/n) + 3√(n)((n + 1)/n) − n > 0, which is absurd. So σP(w1) ≤ 1 − r. The latter implies that bB(w1) = 0, which contradicts bB(w1) = 1. So c1 > 1/n. By similar reasoning concerning permutations of B and P, it follows that for all i: ci > 1/n, which contradicts LW.□
PSFC, along with all instances of SFr, are also incompatible with CL and IP (in the presence of reasonable principles):
Theorem 14
{Ub, Ab, Nb, Zb, Up, PSFC, CL} is inconsistent.
Theorem 15
For all r in (0, 1): {Ub, Ab, Nb, Zb, Up, SFr, CL} is inconsistent.
Theorem 16
{Ub, Ab, Nb, Zb, Up, Ap, Np, PSFC, IP} is inconsistent.
Theorem 17
For all r in (0, 1): {Ub, Ab, Nb, Zb, Up, Ap, Np, SFr, IP} is inconsistent.
The proofs of Theorems 8 and 9 are annotated to indicate the modifications required to prove Theorems 14, 15, 16, and 17.
Appendix B: Proofs
Fact 2.
If Ub, Ab, and S, then for all B, φ: bB(φ) = 1 if and only if for all bi in B: bi(φ) = 1.
Proof
The right to left direction of the consequent follows from the consistency requirement on bB. To establish the left to right direction, it is sufficient to show that for all n, m: if 0 ≤ m < n, then there exists some sets of belief functions B, such that \( n_{1j} \) + … + \( n_{nj} \) = n − m, and \( n_{j} \) = 1. If there are such sets of belief functions, then we have, for all m, such that 0 ≤ m < n, a proposition ¬wj, that is believed by m of n agents, but not by the collective. It is straightforward to show that all of the needed B exist. By Ub, we have for all n and m, such that 0 ≤ m < n, some B such that the columns of B are the set of permutations of m 0 s and n − m 1 s. For each such B, \( n_{i} \) = 1, for all \( n_{i} \) in bB, given Ab and Nb. The preceding holds, since, for each such B, each \( n_{i} \) must take the same value, given Ab and Nb, which must be 1, by the consistency requirement on bB.□
Theorem 1
For all r: if 0.5 < r < 1, then {Ub, A b, Nb, Up, Ap, Np, Lr} is inconsistent.
Proof
Notice that for all r: if 0.5 < r < 1, then there exists n and ε: 0 < ε < 1/n and r = ((n − 1)/n) + ε. In light of the preceding, consider the instances of the following schema, for all n and ε: 0 < ε < 1/n. Let Π = {w1, …, wn}. Let Bn = 〈b1, …, bn〉, where bi has the value 0 in the ith position, and the value 1 in all other positions. Let Pn = 〈p1, …, pn〉, where pi has the value (1/n) − ε in the ith position, and the value (1/n) + (ε/(n − 1)) in all other positions. In this case, Ab and Nb imply that bBn has the value 1 in every position, and Ap and Np imply pPn has the value 1/n in every position. Notice that for all i: 〈bi, pi〉 is Lockean at ((n − 1)/n) + ε, but 〈bBn, pPn〉 is not Lockean at ((n − 1)/n) + ε.□
Theorem 3
For all F: if {F = Fbel, Ub, Ab, Nb} is consistent, then {F = Fbel, Up, Ap, Np, PLC} is consistent.
Proof
To demonstrate the result, we show how to define a function Fprob that satisfies Up, Ap, Np, and PLC, given any aggregation function Fbel that satisfies Ab and Nb. Given any P, we define B* = 〈b*1,…, b*n〉 to be the (unique) set of belief functions, such that for all i: CL(b*i, pi) [or such that CSF(b*i, pi) or SFr(b*i, pi) for the proofs of Theorems 12 and 13, respectively]. We then define Fprob by \( r_{j} \) = 1/(\( n^{*}_{1} \) +···+ \( n^{*}_{k} \)), if \( n^{*}_{j} \) = 1, and \( r_{j} \) = 0, if \( n^{*}_{j} \) = 0 (where \( n^{*}_{1} \) through \( n^{*}_{k} \) are determined via Fbel, given Ub). [Notice that, in the preceding step, we require that |{ \( n^{*}_{j} \) | \( n^{*}_{j} \) = 1}| is finite, which holds if Π is finite, but also assuming Zb, since all p-stable sets are finite (Leitgeb 2013, p. 1366).] It follows immediately from the definition of Fprob that the combination of Fbel and Fprob satisfy PLC [or PSFC or SFr, respectively]. The inputs to Fprob are not restricted; so Up is also satisfied. Ap requires that for all P and g: if g: {1, …, n} → {1, …, n} is a permutation, and P′ = 〈pg(1), …, pg(n)〉, then pP′ = pP. We assume for arbitrary P and g, that g: {1, …, n} → {1, …, n} is a permutation, and P′ = 〈pg(1), …, pg(n)〉. We show that pP′ = pP. To begin with, notice that P determines B*, and P′ determines B′*, so that b′*i = b*g(i). But Ab, so B′* = B*, and thus pP′ = pP, given the definition of Fprob. So Ap is satisfied. Np requires that for all P: if f: {1, …, k} → {1, …, k} is a permutation, P′ = 〈p′1, …, p′n〉, and for all i: p′i = 〈\( r_{if\left( 1 \right)} \), …, \( r_{if\left( k \right)} \)〉, then pP′ = 〈\( r_{f\left( 1 \right)} \), …, \( r_{f\left( k \right)} \)〉. We assume for arbitrary P and f, that f: {1, …, k} → {1, …, k} is a permutation, P′ = 〈p′1, …, p′n〉, and for all i: p′i = 〈\( r_{if\left( 1 \right)} \), …, \( r_{if\left( k \right)} \)〉. We show that pP′ = 〈\( r_{f\left( 1 \right)} \), …, \( r_{f\left( k \right)} \)〉. To begin with, notice that P determines B*, and P′ determines B′*, so that b′*i = b*g(i). But Ab, so B′* = B*, and thus pP′ = pP, given the definition of Fprob. So Ap is satisfied. Np requires that for all P: if f: {1, …, k} → {1, …, k} is a permutation, P′ = 〈p′1, …, p′n〉, and for all i: p′i = 〈\( r_{if\left( 1 \right)} \), …, \( r_{if\left( k \right)} \)〉, then pP′ = 〈\( r_{f\left( 1 \right)} \), …, \( r_{f\left( k \right)} \)〉. We assume for arbitrary P and f, that f: {1, …, k} → {1, …, k} is a permutation, P′ = 〈p′1, …, p′n〉, and for all i: p′i = 〈\( r_{if\left( 1 \right)} \), …, \( r_{if\left( k \right)} \)〉. We show that pP′ = 〈\( r_{f\left( 1 \right)} \), …, \( r_{f\left( k \right)} \)〉. To begin with, notice that P determines B*, and P′ determines B′*, such that b*i = 〈\( n_{i1} \), …, \( n_{ik} \)〉 and b′*i = 〈\( n_{if\left( 1 \right)} \), …, \( n_{if\left( k \right)} \)〉. But Nb, so pB′* = 〈\( n_{f\left( 1 \right)} \), …, \( n_{f\left( k \right)} \)〉. So pP′ = 〈\( r_{f\left( 1 \right)} \), …, \( r_{f\left( k \right)} \)〉, given Fprob. So Np is satisfied.□
Theorem 4
{Ub, Ab, Nb, Up, LW, PLC} is inconsistent.
Proof
Let Π = {w1, w2}. Consider B1 = 〈〈1, 0〉, 〈0, 1〉〉, and P1 = 〈〈0.51, 0.49〉, 〈0.01, 0.99〉〉. Next consider B2 = 〈〈1, 0〉, 〈0, 1〉〉, and P2 = 〈〈0.99, 0.01〉, 〈0.49, 0.51〉〉. Notice that Ab and Nb imply that bB1 = bB2 = 〈1, 1〉. Notice that PLC (and similarly PSFC) imply that pP1 = 〈0.5, 0.5〉, which, according LW, holds only if 0.51·c1 + 0.01·c2 = 0.5, and so where c1 = 0.98 and c2 = 0.02. But PLC (and similarly PSFC) also imply that pP2 = 〈0.5, 0.5〉, which, according LW, holds only if 0.01·c1 + 0.51·c2 = 0.5, and so where c1 = 0.02 and c2 = 0.98, which is a contradiction.□
Theorem 5
For all F: if {F = Fbel, Ub, Ab, Nb, UNb} is consistent, then {F = Fbel, Up, Ap, Np, UNp, PLC} is consistent.
Proof
The proof proceeds as the proof of Theorem 5, save that we make the further assumption that UNb, and show that Fprob satisfies UNp. Assume not. Then there is a case where (i) \( r_{j} \) ≠ 0 and for all i: \( r_{ij} \) = 0, or a case where (ii) \( r_{j} \) ≠ 1 and for all i: \( r_{ij} \) = 1. In case (i), we have for all i: \( n^{*}_{ij} \) = 0, given for all i: \( r_{ij} \) = 0 (by the definition of B*). So \( n^{*}_{j} \) = 0, given UNb. And so \( r_{j} \) = 0, by the definition of Fprob. In case (ii), we have for all i: \( n^{*}_{ij} \) = 1 and for all i, k ≠ j: \( n^{*}_{ik} \) = 0, given for all i: \( r_{ij} \) = 1 (by the definition of B*). So \( n^{*}_{j} \) = 1, and for all k ≠ j: \( n^{*}_{k} \) = 0, given Ub and UNb. So \( r_{j} \) = 1, by the definition of Fprob.□
Theorem 6
For all F: if {F = Fbel, Ub, Ab, Nb, WDb} is consistent, then {F = Fbel, Up, Ap, Np, WDp, PLC} is consistent.
Proof
The proof proceeds as the proof of Theorem 5, save that we make the further assumption that WDb, and show that Fprob satisfies WDp. Let P, j, and k be arbitrary, with for all i: \( r_{ij} \) ≥ \( r_{ik} \). In that case, for all i: \( n^{*}_{ij} \)≥ \( n^{*}_{ik} \), by the definition of Fprob. So \( n^{*}_{j} \)≥ \( n^{*}_{k} \), given WDb. So \( r_{j} \) ≥ \( r_{k} \), by the definition of Fprob.□
Theorem 8
{Ub, Ab, Nb, Zb, Up, PLC, CL} is inconsistent.
Proof
Let Π = {w1, w2, w3}. Consider B = 〈〈1, 0, 0〉, 〈0, 0, 1〉〉, and P = 〈〈0.9, 0.09, 0.01〉, 〈0.01, 0.09, 0.9〉〉. Notice that Ab, Nb, and Zb imply that bB = 〈1, 0, 1〉. Without loss of generality, assume that pp = 〈a, b, c〉. Then PLC implies that a > b, since CL(〈1, 0, 0〉, 〈0.9, 0.09, 0.01〉) and CL(〈0, 0, 1〉, 〈0.01, 0.09, 0.99〉) [and similarly CSF(〈1, 0, 0〉, 〈0.9, 0.09, 0.01〉) and CSF(〈0, 0, 1〉, 〈0.01, 0.09, 0.99〉)Footnote 23]. Now consider P′ = 〈〈90/99, 9/99, 0〉, 〈1/10, 9/10, 0〉〉, and B′ = 〈〈1, 0, 0〉, 〈0, 1, 0〉〉. Notice that Ab, Nb, and Zb imply that bB′ = 〈1, 1, 0〉. Now notice that CL implies that pp′ = 〈a/(a + b), b/(a + b), 0〉. Finally, notice that CL(〈1, 0, 0〉, 〈90/99, 9/99, 0〉) and CL(〈0, 1, 0〉, 〈1/10, 9/10, 0〉) imply that CL(〈1, 1, 0〉, 〈a/(a + b), b/(a + b), 0〉), which is absurd, since CL(〈1, 0, 0〉, 〈a/(a + b), b/(a + b), 0〉), given a > b [and similarly with CSF in place of CL].□
Theorem 9
{Ub, Ab, Nb, Zb, Up, Ap, Np, PLC, IP} is inconsistent.
Proof
Let Π = {w1, w2, w3, w4}. Consider B = 〈〈0, 0, 0, 1〉, 〈1, 0, 0, 0〉〉, and P = 〈〈0.01, 0.09, 0.09, 0.81〉, 〈0.81, 0.09, 0.09, 0.01〉〉.Footnote 24 Notice that Ab, Nb, and Zb imply that bB = 〈1, 0, 0, 1〉. Now notice that Ap and Np imply that pP(w1) = pP(w4) and pP(w2) = pP(w3), and this implies that pP(w1∪w2) = pP(w1∪w3) = 1/2. Now notice that CL(〈0, 0, 0, 1〉, 〈0.01, 0.09, 0.09, 0.81〉) and CL(〈〈1, 0, 0, 0〉, 〈0.81, 0.09, 0.9, 01〉) [and similarly CSF(〈0, 0, 0, 1〉, 〈0.01, 0.09, 0.09, 0.81〉) and CSF(〈1, 0, 0, 0〉, 〈0.81, 0.09, 0.09, 0.01〉)]. So PLC implies that CL(〈1, 0, 0, 1〉, PP) [and similarly CSF(〈1, 0, 0, 1〉, PP)], which implies that pP(w1) > pP(w2), and thus pP(w1) > 1/4. Finally, notice that p1((w1∪w2)∩(w1∪w3)) = p1(w1∪w2)p1(w1∪w3) and p2((w1∪w2)∩(w1∪w3)) = p2(w1∪w2)p1(w1∪w3). So IP implies that pP((w1∪w2)∩(w1∪w3)) = pP(w1∪w2)pP(w1∪w3). But pP((w1∪w2)∩(w1∪w3)) = pP(w1), and, since pP(w1∪w2) = pP(w1∪w3), pP(w1) = 1/4, which is absurd.□
Rights and permissions
About this article
Cite this article
Thorn, P.D. The joint aggregation of beliefs and degrees of belief. Synthese 197, 5389–5409 (2020). https://doi.org/10.1007/s11229-018-01966-0
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11229-018-01966-0