Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Argument evaluation in multi-agent logics
Alfredo Burrieza and Antonio Yuste-Ginel1 University of M´alaga
Department of Philosophy MBR 2018, Sevilla, October 2018
1
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Outline
Introduction
The logicAE
Syntax Semantics
Assumable properties
Modelling argument evaluation inAE
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Outline
Introduction
The logicAE
Syntax Semantics
Assumable properties
Modelling argument evaluation inAE
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
The general question
(i) Lettbe an argument
is t a good argument?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
The general question
(i) Lettbe an argument
is t a good argument?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
The general question
(i) Lettbe an argument
is t a good argument?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
The general question
(i) Lettbe an argument
is t a good argument?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
The general question
(i) Lettbe an argument
is t a good argument?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
What does it make a set of premises a good one?
What does it make an inference link a good one?
First Thesis:
Argumentative goodness is a context-dependent notion
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
What does it make a set of premises a good one?
What does it make an inference link a good one?
First Thesis:
Argumentative goodness is a context-dependent notion
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
What does it make a set of premises a good one?
What does it make an inference link a good one?
First Thesis:
Argumentative goodness is a context-dependent notion
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
What does it make a set of premises a good one?
What does it make an inference link a good one?
First Thesis:
Argumentative goodness is a context-dependent notion
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
→A good answer should be sensitive to other questions:
� Good for whom? (agent perspective)
� Good for what? (goal of evaluation)
� How good is it compare to other arguments?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
→A good answer should be sensitive to other questions:
� Good for whom? (agent perspective)
� Good for what? (goal of evaluation)
� How good is it compare to other arguments?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
→A good answer should be sensitive to other questions:
� Good for whom? (agent perspective)
� Good for what? (goal of evaluation)
� How good is it compare to other arguments?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
→A good answer should be sensitive to other questions:
� Good for whom? (agent perspective)
� Good for what? (goal of evaluation)
� How good is it compare to other arguments?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
→A good answer should be sensitive to other questions:
� Good for whom? (agent perspective)
� Good for what? (goal of evaluation)
� How good is it compare to other arguments?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
If we combine these three factors we obtain:
(ii) how good ist compare tot�, from a’sperspective if she intends to do Goal?
Some instantiations ofGoal:
� “justifying a given proposition ϕ”
� “deciding if ϕis true”
� “convincing agentb (or group G) thatϕis the case”
� “convincing someone of doing α”
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
If we combine these three factors we obtain:
(ii) how good ist compare tot�, from a’sperspective if she intends to do Goal?
Some instantiations ofGoal:
� “justifying a given proposition ϕ”
� “deciding if ϕis true”
� “convincing agentb (or group G) thatϕis the case”
� “convincing someone of doing α”
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
If we combine these three factors we obtain:
(ii) how good ist compare tot�, from a’sperspective if she intends to do Goal?
Some instantiations ofGoal:
� “justifying a given proposition ϕ”
� “deciding if ϕis true”
� “convincing agentb (or group G) thatϕis the case”
� “convincing someone of doing α”
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
If we combine these three factors we obtain:
(ii) how good ist compare tot�, from a’sperspective if she intends to do Goal?
Some instantiations ofGoal:
� “justifying a given proposition ϕ”
� “deciding if ϕis true”
� “convincing agentb (or group G) thatϕis the case”
� “convincing someone of doing α”
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
If we combine these three factors we obtain:
(ii) how good ist compare tot�, from a’sperspective if she intends to do Goal?
Some instantiations ofGoal:
� “justifying a given proposition ϕ”
� “deciding if ϕis true”
� “convincing agentb (or group G) thatϕis the case”
� “convincing someone of doing α”
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
If we combine these three factors we obtain:
(ii) how good ist compare tot�, from a’sperspective if she intends to do Goal?
Some instantiations ofGoal:
� “justifying a given proposition ϕ”
� “deciding if ϕis true”
� “convincing agentb (or group G) thatϕis the case”
� “convincing someone of doing α”
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
If we combine these three factors we obtain:
(ii) how good ist compare tot�, from a’sperspective if she intends to do Goal?
Some instantiations ofGoal:
� “justifying a given proposition ϕ”
� “deciding if ϕis true”
� “convincing agentb (or group G) thatϕis the case”
� “convincing someone of doing α”
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
If we combine these three factors we obtain:
(ii) how good ist compare tot�, from a’sperspective if she intends to do Goal?
Some instantiations ofGoal:
� “justifying a given proposition ϕ”
� “deciding if ϕis true”
� “convincing agentb (or group G) thatϕis the case”
� “convincing someone of doing α”
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
If we combine these three factors we obtain:
(ii) how good ist compare tot�, from a’sperspective if she intends to do Goal?
Some instantiations ofGoal:
� “justifying a given proposition ϕ”
� “deciding if ϕis true”
� “convincing agentb (or group G) thatϕis the case”
� “convincing someone of doing α”
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
If we combine these three factors we obtain:
(ii) how good ist compare tot�, from a’sperspective if she intends to do Goal?
Some instantiations ofGoal:
� “justifying a given proposition ϕ”
� “deciding if ϕis true”
� “convincing agentb (or group G) thatϕis the case”
� “convincing someone of doing α”
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
If we combine these three factors we obtain:
(ii) how good ist compare tot�, from a’sperspective if she intends to do Goal?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
If we combine these three factors we obtain:
(ii) how good ist compare tot�, from a’sperspective if she intends to do Goal?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Outline
Introduction
The logicAE
Syntax Semantics
Assumable properties
Modelling argument evaluation inAE
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Selecting the tool-kit
� Arguments → Justification Logic (Artemov and Nogina, 2005; Artemov, 2008, 2012; Baltag et al., 2012, 2014)
� Doxastic attitudes of evaluative agents→Doxastic logic(Fagin et al., 2004)
� Expert opinions on different topics→ Logics for Belief Dependence(Huang, 1990)
� Availability of information → Awareness Logic(Fagin and Halpern, 1987; Baltag et al., 2012)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Selecting the tool-kit
� Arguments→ Justification Logic (Artemov and Nogina, 2005; Artemov, 2008, 2012; Baltag et al., 2012, 2014)
� Doxastic attitudes of evaluative agents→Doxastic logic(Fagin et al., 2004)
� Expert opinions on different topics→ Logics for Belief Dependence(Huang, 1990)
� Availability of information → Awareness Logic(Fagin and Halpern, 1987; Baltag et al., 2012)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Selecting the tool-kit
� Arguments → Justification Logic (Artemov and Nogina, 2005; Artemov, 2008, 2012; Baltag et al., 2012, 2014)
� Doxastic attitudes of evaluative agents→Doxastic logic(Fagin et al., 2004)
� Expert opinions on different topics→ Logics for Belief Dependence(Huang, 1990)
� Availability of information → Awareness Logic(Fagin and Halpern, 1987; Baltag et al., 2012)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Selecting the tool-kit
� Arguments → Justification Logic (Artemov and Nogina, 2005; Artemov, 2008, 2012; Baltag et al., 2012, 2014)
� Doxastic attitudes of evaluative agents
→Doxastic logic(Fagin et al., 2004)
� Expert opinions on different topics→ Logics for Belief Dependence(Huang, 1990)
� Availability of information → Awareness Logic(Fagin and Halpern, 1987; Baltag et al., 2012)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Selecting the tool-kit
� Arguments → Justification Logic (Artemov and Nogina, 2005; Artemov, 2008, 2012; Baltag et al., 2012, 2014)
� Doxastic attitudes of evaluative agents→Doxastic logic(Fagin et al., 2004)
� Expert opinions on different topics→ Logics for Belief Dependence(Huang, 1990)
� Availability of information → Awareness Logic(Fagin and Halpern, 1987; Baltag et al., 2012)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Selecting the tool-kit
� Arguments → Justification Logic (Artemov and Nogina, 2005; Artemov, 2008, 2012; Baltag et al., 2012, 2014)
� Doxastic attitudes of evaluative agents→Doxastic logic(Fagin et al., 2004)
� Expert opinions on different topics
→ Logics for Belief Dependence(Huang, 1990)
� Availability of information → Awareness Logic(Fagin and Halpern, 1987; Baltag et al., 2012)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Selecting the tool-kit
� Arguments → Justification Logic (Artemov and Nogina, 2005; Artemov, 2008, 2012; Baltag et al., 2012, 2014)
� Doxastic attitudes of evaluative agents→Doxastic logic(Fagin et al., 2004)
� Expert opinions on different topics→ Logics for Belief Dependence(Huang, 1990)
� Availability of information → Awareness Logic(Fagin and Halpern, 1987; Baltag et al., 2012)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Selecting the tool-kit
� Arguments → Justification Logic (Artemov and Nogina, 2005; Artemov, 2008, 2012; Baltag et al., 2012, 2014)
� Doxastic attitudes of evaluative agents→Doxastic logic(Fagin et al., 2004)
� Expert opinions on different topics→ Logics for Belief Dependence(Huang, 1990)
� Availability of information
→ Awareness Logic(Fagin and Halpern, 1987; Baltag et al., 2012)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Selecting the tool-kit
� Arguments → Justification Logic (Artemov and Nogina, 2005; Artemov, 2008, 2012; Baltag et al., 2012, 2014)
� Doxastic attitudes of evaluative agents→Doxastic logic(Fagin et al., 2004)
� Expert opinions on different topics→ Logics for Belief Dependence(Huang, 1990)
� Availability of information → Awareness Logic(Fagin and Halpern, 1987; Baltag et al., 2012)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Selecting the tool-kit
� Arguments → Justification Logic (Artemov and Nogina, 2005; Artemov, 2008, 2012; Baltag et al., 2012, 2014)
� Doxastic attitudes of evaluative agents→Doxastic logic(Fagin et al., 2004)
� Expert opinions on different topics→ Logics for Belief Dependence(Huang, 1990)
� Availability of information → Awareness Logic(Fagin and Halpern, 1987; Baltag et al., 2012)
� Complexity of arguments
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Selecting the tool-kit
� Arguments → Justification Logic (Artemov and Nogina, 2005; Artemov, 2008, 2012; Baltag et al., 2012, 2014)
� Doxastic attitudes of evaluative agents→Doxastic logic(Fagin et al., 2004)
� Expert opinions on different topics→ Logics for Belief Dependence(Huang, 1990)
� Availability of information → Awareness Logic(Fagin and Halpern, 1987; Baltag et al., 2012)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Outline
Introduction
The logicAE
Syntax
Semantics
Assumable properties
Modelling argument evaluation inAE
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Language
P (atoms),A (agents); termsT and formulasF are defined:
t::=cϕ |[t+t]|[t·t]
ϕ::=p| ¬ϕ|(ϕ∧ϕ)|t�ϕ|Baϕ|Dabϕ|Eat|Com≤(t, s)
Terms intend to represent arguments
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Language
P (atoms),A (agents);
termsT and formulasF are defined:
t::=cϕ |[t+t]|[t·t]
ϕ::=p| ¬ϕ|(ϕ∧ϕ)|t�ϕ|Baϕ|Dabϕ|Eat|Com≤(t, s)
Terms intend to represent arguments
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Language
P (atoms),A (agents); termsT and formulasF are defined:
t::=cϕ |[t+t]|[t·t]
ϕ::=p| ¬ϕ|(ϕ∧ϕ)|t�ϕ|Baϕ|Dabϕ|Eat|Com≤(t, s)
Terms intend to represent arguments
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Language
P (atoms),A (agents); termsT and formulasF are defined:
t::=cϕ |[t+t]|[t·t]
ϕ::=p| ¬ϕ|(ϕ∧ϕ)|t�ϕ|Baϕ|Dabϕ|Eat|Com≤(t, s)
Terms intend to represent arguments
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Language
P (atoms),A (agents); termsT and formulasF are defined:
t::=cϕ |[t+t]|[t·t]
ϕ::=p| ¬ϕ|(ϕ∧ϕ)|t�ϕ|Baϕ|Dabϕ|Eat|Com≤(t, s)
Terms intend to represent arguments
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Language
P (atoms),A (agents); termsT and formulasF are defined:
t::=cϕ |[t+t]|[t·t]
ϕ::=p| ¬ϕ|(ϕ∧ϕ)|t�ϕ|Baϕ|Dabϕ|Eat|Com≤(t, s)
Terms intend to represent arguments
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Language
P (atoms),A (agents); termsT and formulasF are defined:
t::=cϕ |[t+t]|[t·t]
ϕ::=p| ¬ϕ|(ϕ∧ϕ)|t�ϕ|Baϕ|Dabϕ|Eat|Com≤(t, s)
Terms intend to represent arguments
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Language
P (atoms),A (agents); termsT and formulasF are defined:
t::=cϕ |[t+t]|[t·t]
ϕ::=p| ¬ϕ|(ϕ∧ϕ)|t�ϕ|Baϕ|Dabϕ|Eat|Com≤(t, s)
Terms intend to represent arguments
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Arguments
t::=cϕ |[t+t]|[t·t]
Taken from (Baltag et al., 2012, 2014).
cϕ:≈“a premise assertingϕ”
t+s:≈“result from combining tand swithout performing any inference”
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Arguments
t::=cϕ |[t+t]|[t·t]
Taken from (Baltag et al., 2012, 2014).
cϕ:≈“a premise assertingϕ”
t+s:≈“result from combining tand swithout performing any inference”
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Arguments
t::=cϕ |[t+t]|[t·t]
Taken from (Baltag et al., 2012, 2014).
cϕ:≈“a premise assertingϕ”
t+s:≈“result from combining tand swithout performing any inference”
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Arguments
t::=cϕ |[t+t]|[t·t]
Taken from (Baltag et al., 2012, 2014).
cϕ:≈“a premise assertingϕ”
t+s:≈“result from combiningt and swithout performing any inference”
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Arguments
t::=cϕ |[t+t]|[t·t]
Taken from (Baltag et al., 2012, 2014).
cϕ:≈“a premise assertingϕ”
t+s:≈“result from combiningt and swithout performing any inference”
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Arguments
t::=cϕ |[t+t]|[t·t]
Taken from (Baltag et al., 2012, 2014).
cϕ:≈“a premise assertingϕ”
t+s:≈“result from combiningt and swithout performing any inference”
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Arguments
t::=cϕ |[t+t]|[t·t]
Note: other operations between terms are definable without any technical problem (as long as they preserve truth)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Arguments
t::=cϕ |[t+t]|[t·t]
Note: other operations between terms are definable without any technical problem (as long as they preserve truth)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Arguments
t::=cϕ |[t+t]|[t·t]
Note: other operations between terms are definable without any technical problem (as long as they preserve truth)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Formulas
ϕ::=p| ¬ϕ|(ϕ∧ϕ)|t�ϕ|Baϕ|Dabϕ|Eat|Com≤(t, s)
Baϕ:≈“ agent abelieves thatϕ”
Dabϕ:≈“b is an experton topicϕ according toa’s opinion”
t�ϕ:≈“argument thas propositionϕ as itsconclusion”
Com≤(t, s) :≈“argument tisat least as simple as arguments”
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Formulas
ϕ::=p| ¬ϕ|(ϕ∧ϕ)|t�ϕ|Baϕ|Dabϕ|Eat|Com≤(t, s)
Baϕ:≈“ agent abelieves thatϕ”
Dabϕ:≈“b is an experton topicϕ according toa’s opinion”
t�ϕ:≈“argument thas propositionϕ as itsconclusion”
Com≤(t, s) :≈“argument tisat least as simple as arguments”
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Formulas
ϕ::=p| ¬ϕ|(ϕ∧ϕ)|t�ϕ|Baϕ|Dabϕ|Eat|Com≤(t, s)
Baϕ:≈“ agent abelieves thatϕ”
Dabϕ:≈“b is an experton topicϕ according toa’s opinion”
t�ϕ:≈“argument thas propositionϕ as itsconclusion”
Com≤(t, s) :≈“argument tisat least as simple as arguments”
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Formulas
ϕ::=p| ¬ϕ|(ϕ∧ϕ)|t�ϕ|Baϕ|Dabϕ|Eat|Com≤(t, s)
Baϕ:≈“ agent abelieves thatϕ”
Dabϕ:≈“b is an experton topicϕ according toa’s opinion”
t�ϕ:≈“argument thas propositionϕas its conclusion”
Com≤(t, s) :≈“argument tisat least as simple as arguments”
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Formulas
ϕ::=p| ¬ϕ|(ϕ∧ϕ)|t�ϕ|Baϕ|Dabϕ|Eat|Com≤(t, s)
Baϕ:≈“ agent abelieves thatϕ”
Dabϕ:≈“b is an experton topicϕ according toa’s opinion”
t�ϕ:≈“argument thas propositionϕas its conclusion”
Com≤(t, s) :≈“argument tisat least as simple as arguments”
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Syntax
Formulas
ϕ::=p| ¬ϕ|(ϕ∧ϕ)|t�ϕ|Baϕ|Dabϕ|Eat|Com≤(t, s)
Baϕ:≈“ agent abelieves thatϕ”
Dabϕ:≈“b is an experton topicϕ according toa’s opinion”
t�ϕ:≈“argument thas propositionϕas its conclusion”
Com≤(t, s) :≈“argument tisat least as simple as arguments”
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Outline
Introduction
The logicAE
Syntax
Semantics
Assumable properties
Modelling argument evaluation inAE
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Interpreting “syntactic” formulas outside the model
Define�⊆ T × T as the smallest set such that:
� cϕ �ϕ
� If t�ϕ→ψ and s�ϕ, then [t·s]�ψ
� If t�ϕand s�ψ, then [t+s]�ϕ
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Interpreting “syntactic” formulas outside the model
Define�⊆ T × T as the smallest set such that:
� cϕ �ϕ
� If t�ϕ→ψ and s�ϕ, then [t·s]�ψ
� If t�ϕand s�ψ, then [t+s]�ϕ
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Interpreting “syntactic” formulas outside the model
Define�⊆ T × T as the smallest set such that:
� cϕ �ϕ
� If t�ϕ→ψ and s�ϕ, then [t·s]�ψ
� If t�ϕand s�ψ, then [t+s]�ϕ
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Interpreting “syntactic” formulas outside the model
Define�⊆ T × T as the smallest set such that:
� cϕ �ϕ
� If t�ϕ→ψ and s�ϕ, then [t·s]�ψ
� If t�ϕand s�ψ, then [t+s]�ϕ
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Interpreting “syntactic” formulas outside the model
Define�⊆ T × T as the smallest set such that:
� cϕ �ϕ
� If t�ϕ→ψ and s�ϕ, then [t·s]�ψ
� If t�ϕand s�ψ, then [t+s]�ϕ
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Interpreting “syntactic” formulas outside the model
Define�⊆ T × T as the smallest set such that:
� cϕ �ϕ
� If t�ϕ→ψ and s�ϕ, then [t·s]�ψ
� If t�ϕand s�ψ, then [t+s]�ϕ
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Interpreting “syntactic” formulas outside the model
Define�⊆ T × T as the smallest set such that:
� cϕ �ϕ
� If t�ϕ→ψ and s�ϕ, then [t·s]�ψ
� If t�ϕand s�ψ, then [t+s]�ϕ
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Interpreting “syntactic” formulas outside the model
Define�⊆ T × T as the smallest set such that:
� cϕ �ϕ
� If t�ϕ→ψ and s�ϕ, then [t·s]�ψ
� If t�ϕand s�ψ, then [t+s]�ϕ
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Interpreting “syntactic” formulas outside the model
Define�⊆ T × T as the smallest set such that:
� cϕ �ϕ
� If t�ϕ→ψ and s�ϕ, then [t·s]�ψ
� If t�ϕand s�ψ, then [t+s]�ϕ
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Interpreting “syntactic” formulas outside the model
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Interpreting “syntactic” formulas outside the model
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Models and truth
M:=�W,{Ra},{Da},{Ea}, V�
for eacha∈ A
� W �=∅ (possible worlds)
� Ra⊆W ×W (doxastic relation)
� Da: (A ×W)−→℘(F) (advisor function)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Models and truth
M:=�W,{Ra},{Da},{Ea}, V�
for eacha∈ A
� W �=∅ (possible worlds)
� Ra⊆W ×W (doxastic relation)
� Da: (A ×W)−→℘(F) (advisor function)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Models and truth
M:=�W,{Ra},{Da},{Ea}, V�
for eacha∈ A
� W �=∅ (possible worlds)
� Ra⊆W ×W (doxastic relation)
� Da: (A ×W)−→℘(F) (advisor function)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Models and truth
M:=�W,{Ra},{Da},{Ea}, V�
for eacha∈ A
� W �=∅ (possible worlds)
� Ra⊆W ×W (doxastic relation)
� Da: (A ×W)−→℘(F) (advisor function)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Models and truth
M:=�W,{Ra},{Da},{Ea}, V�
for eacha∈ A
� W �=∅ (possible worlds)
� Ra⊆W ×W (doxastic relation)
� Da: (A ×W)−→℘(F) (advisor function)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Models and truth
M:=�W, Ra,Da,Ea, V�a∈A
M, w�t�ϕ iff ϕ∈conclusions(t)
M, w�Com≤(t, s) iff length(t) +|atom(t)| ≤
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Models and truth
M:=�W, Ra,Da,Ea, V�a∈A
M, w�t�ϕ iff ϕ∈conclusions(t)
M, w�Com≤(t, s) iff length(t) +|atom(t)| ≤
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Models and truth
M:=�W, Ra,Da,Ea, V�a∈A
M, w�t�ϕ iff ϕ∈conclusions(t)
M, w�Com≤(t, s) iff length(t) +|atom(t)| ≤
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Models and truth
M:=�W, Ra,Da,Ea, V�a∈A
M, w�t�ϕ iff ϕ∈conclusions(t)
M, w�Com≤(t, s) iff length(t) +|atom(t)| ≤
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Models and truth
M:=�W, Ra,Da,Ea, V�a∈A
M, w�Baϕ iff ∀w�(wRaw� ⇒ M, w� �ϕ)
M, w�Dabϕ iff ϕ∈ Da(b, w)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Models and truth
M:=�W, Ra,Da,Ea, V�a∈A
M, w�Baϕ iff ∀w�(wRaw� ⇒ M, w� �ϕ)
M, w�Dabϕ iff ϕ∈ Da(b, w)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Models and truth
M:=�W, Ra,Da,Ea, V�a∈A
M, w�Baϕ iff ∀w�(wRaw� ⇒ M, w� �ϕ)
M, w�Dabϕ iff ϕ∈ Da(b, w)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Semantics
Models and truth
M:=�W, Ra,Da,Ea, V�a∈A
M, w�Baϕ iff ∀w�(wRaw� ⇒ M, w� �ϕ)
M, w�Dabϕ iff ϕ∈ Da(b, w)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Assumable properties
Outline
Introduction
The logicAE
Syntax Semantics
Assumable properties
Modelling argument evaluation inAE
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Assumable properties
Minimal system
The minimal logic for argument evaluationLmin
Axioms
All propositional tautologies
�t�ϕwhenevert�ϕ
� ¬(t�ϕ) whenever nott�ϕ (admissibility)
�Ba(ϕ→ψ)→(Baϕ→Baψ) (normality ofBa)
� Com≤(t, s) whenever length(t) +|atom(t)| ≤ length(s) +
|atom(s)|
� ¬Com≤(t, s) wheneverlength(t) +|atom(t)| �length(s) +
|atom(s)| Rules
M P �ϕ,�ϕ→ψ⇒ �ψ
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Assumable properties
Assumable properties
Name Axiom scheme
Consistency of beliefs � ¬Ba⊥
Positive introspection �Baϕ→BaBaϕ
Negative introspection � ¬Baϕ→Ba¬Baϕ
Neutrality ofDab �Dab¬ϕ→Dabϕ
�Dabϕ→Dab¬ϕwhereϕ�=¬ψ
Safety ofDab �Dabϕ↔BaDabϕ
Uniqueness ofDab �Dabϕ→ ¬Dacϕfor eachb�=c
Transparency of awareness �Eat↔BaEat
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Assumable properties
Assumable properties
Name Axiom scheme
Consistency of beliefs � ¬Ba⊥
Positive introspection �Baϕ→BaBaϕ
Negative introspection � ¬Baϕ→Ba¬Baϕ
Neutrality ofDab �Dab¬ϕ→Dabϕ
�Dabϕ→Dab¬ϕwhereϕ�=¬ψ
Safety ofDab �Dabϕ↔BaDabϕ
Uniqueness ofDab �Dabϕ→ ¬Dacϕfor eachb�=c
Transparency of awareness �Eat↔BaEat
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Assumable properties
Theorem (Soundness, strong completeness and decidability)
Any extensionL• of Lmin with any set of schemes in the table
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Introduction
The logicAE
Syntax Semantics
Assumable properties
Modelling argument evaluation inAE
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Evaluation and preferences
Note: The result of arguments evaluations can be understood preference relations over arguments, for each agent.
� We can try to establish these relations by the use of other notions expressible in our language.
� Different preference relations should be defined in the object language for different instantiations of:
(ii) how good ist compare tot�, from i’sperspective if she intends to doGoal?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Evaluation and preferences
Note: The result of arguments evaluations can be understood preference relations over arguments, for each agent.
� We can try to establish these relations by the use of other notions expressible in our language.
� Different preference relations should be defined in the object language for different instantiations of:
(ii) how good ist compare tot�, from i’sperspective if she intends to doGoal?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Evaluation and preferences
Note: The result of arguments evaluations can be understood preference relations over arguments, for each agent.
� We can try to establish these relations by the use of other notions expressible in our language.
� Different preference relations should be defined in the object language for different instantiations of:
(ii) how good ist compare tot�, from i’sperspective if she intends to doGoal?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Evaluation and preferences
Note: The result of arguments evaluations can be understood preference relations over arguments, for each agent.
� We can try to establish these relations by the use of other notions expressible in our language.
� Different preference relations should be defined in the object language for different instantiations of:
(ii) how good ist compare tot�, from i’sperspective if she intends to doGoal?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Evaluation and preferences
Note: The result of arguments evaluations can be understood preference relations over arguments, for each agent.
� We can try to establish these relations by the use of other notions expressible in our language.
� Different preference relations should be defined in the object language for different instantiations of:
(ii) how good ist compare tot�, from i’sperspective if she intends to doGoal?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Searching the best argument for
ϕ
How to order the different criteria
Tentative order
A.Structural accuracy: have tand sthe proper syntactic shape?
B.Doxastic acceptance: which premises are better regarding
a’s beliefs?
C.Expert opinion: which premises are better according to the agent thataconsiders an expert on topic ϕ?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Searching the best argument for
ϕ
How to order the different criteria
Tentative order
A.Structural accuracy: have tand sthe proper syntactic shape?
B.Doxastic acceptance: which premises are better regarding
a’s beliefs?
C.Expert opinion: which premises are better according to the agent thataconsiders an expert on topic ϕ?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Searching the best argument for
ϕ
How to order the different criteria
Tentative order
A.Structural accuracy: have tand sthe proper syntactic shape?
B.Doxastic acceptance: which premises are better regarding
a’s beliefs?
C.Expert opinion: which premises are better according to the agent thataconsiders an expert on topic ϕ?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Searching the best argument for
ϕ
How to order the different criteria
Tentative order
A.Structural accuracy: have tand sthe proper syntactic shape?
B.Doxastic acceptance: which premises are better regarding
a’s beliefs?
C.Expert opinion: which premises are better according to the agent thataconsiders an expert on topic ϕ?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Searching the best argument for
ϕ
How to order the different criteria
Tentative order
A.Structural accuracy: have tand sthe proper syntactic shape?
B.Doxastic acceptance: which premises are better regarding
a’s beliefs?
C.Expert opinion: which premises are better according to the agent thataconsiders an expert on topic ϕ?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Searching the best argument for
ϕ
How to order the different criteria
Tentative order
A.Structural accuracy: have tand sthe proper syntactic shape?
B.Doxastic acceptance: which premises are better regarding
a’s beliefs?
C.Expert opinion: which premises are better according to the agent thataconsiders an expert on topic ϕ?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Searching the best argument for
ϕ
How to order the different criteria
Tentative order
A.Structural accuracy: have tand sthe proper syntactic shape?
B.Doxastic acceptance: which premises are better regarding
a’s beliefs?
C.Expert opinion: which premises are better according to the agent thataconsiders an expert on topic ϕ?
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Intuitive principles
A.Structural accuracy:
Ifthas the proper structural shape (t�ϕ) but sdoes not have it (¬(s�ϕ)), thent should be consider strictly better
Pa>(t, s, ϕ)
If bothtand shave defective structures (¬t�ϕ∧ ¬s�ϕ) they should be taken as equally goodPa≈(t, s, ϕ).
If bothtand shave the proper structure (t�ϕ∧s�ϕ), agentashould go on her evaluation to decide which is better.
ρ1 := (t�ϕ∧s�ϕ) :≈“both arguments satisfy structural
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Intuitive principles
A.Structural accuracy:
Ifthas the proper structural shape (t�ϕ) but sdoes not have it (¬(s�ϕ)), thent should be consider strictly better
Pa>(t, s, ϕ)
If bothtand shave defective structures (¬t�ϕ∧ ¬s�ϕ) they should be taken as equally goodPa≈(t, s, ϕ).
If bothtand shave the proper structure (t�ϕ∧s�ϕ), agentashould go on her evaluation to decide which is better.
ρ1 := (t�ϕ∧s�ϕ) :≈“both arguments satisfy structural
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Intuitive principles
A.Structural accuracy:
Ifthas the proper structural shape (t�ϕ) but sdoes not have it (¬(s�ϕ)), thent should be consider strictly better
Pa>(t, s, ϕ)
If bothtand shave defective structures (¬t�ϕ∧ ¬s�ϕ) they should be taken as equally goodPa≈(t, s, ϕ).
If bothtand shave the proper structure (t�ϕ∧s�ϕ), agentashould go on her evaluation to decide which is better.
ρ1 := (t�ϕ∧s�ϕ) :≈“both arguments satisfy structural
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Intuitive principles
A.Structural accuracy:
Ifthas the proper structural shape (t�ϕ) but sdoes not have it (¬(s�ϕ)), thent should be consider strictly better
Pa>(t, s, ϕ)
If bothtand shave defective structures (¬t�ϕ∧ ¬s�ϕ) they should be taken as equally goodPa≈(t, s, ϕ).
If bothtand shave the proper structure (t�ϕ∧s�ϕ), agentashould go on her evaluation to decide which is better.
ρ1 := (t�ϕ∧s�ϕ) :≈“both arguments satisfy structural
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Intuitive principles
A.Structural accuracy:
Ifthas the proper structural shape (t�ϕ) but sdoes not have it (¬(s�ϕ)), thent should be consider strictly better
Pa>(t, s, ϕ)
If bothtand shave defective structures (¬t�ϕ∧ ¬s�ϕ) they should be taken as equally goodPa≈(t, s, ϕ).
If bothtand shave the proper structure (t�ϕ∧s�ϕ), agentashould go on her evaluation to decide which is better.
ρ1 := (t�ϕ∧s�ϕ) :≈“both arguments satisfy structural
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Defining acceptance
B.Doxastic acceptance:
Aat:≈ “aaccepts argumentt, i.e., she believes that all its
premises are true” Baltag et al. (2012)
Rat:≈“ arejects argument t, i.e., she believes that some of its
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Defining acceptance
B.Doxastic acceptance:
Aat:≈“aaccepts argumentt, i.e., she believes that all its
premises are true” Baltag et al. (2012)
Rat:≈“ arejects argument t, i.e., she believes that some of its
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Defining acceptance
B.Doxastic acceptance:
Aat:≈“aaccepts argumentt, i.e., she believes that all its
premises are true” Baltag et al. (2012)
Rat:≈“ arejects argument t, i.e., she believes that some of its
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Defining acceptance
B.Doxastic acceptance:
Aat:≈“aaccepts argumentt, i.e., she believes that all its
premises are true” Baltag et al. (2012)
Rat:≈“ arejects argument t, i.e., she believes that some of its
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Defining acceptance
B.Doxastic acceptance:
Aat:=�cϕ∈sub(t)Baϕ Rat:�cϕ∈sub(t)Ba¬ϕ
Fact: �Rat→ ¬Aatbut�¬Aat→Rat
A>a(t, s) :≈“a considerststrictly more acceptable thans” := (Aat∧ ¬Aas)∨(¬Rat∧Ras)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Defining acceptance
B.Doxastic acceptance:
Aat:=�cϕ∈sub(t)Baϕ
Rat:�cϕ∈sub(t)Ba¬ϕ
Fact: �Rat→ ¬Aatbut�¬Aat→Rat
A>a(t, s) :≈“a considerststrictly more acceptable thans” := (Aat∧ ¬Aas)∨(¬Rat∧Ras)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Defining acceptance
B.Doxastic acceptance:
Aat:=�cϕ∈sub(t)Baϕ Rat:�cϕ∈sub(t)Ba¬ϕ
Fact: �Rat→ ¬Aatbut�¬Aat→Rat
A>a(t, s) :≈“a considerststrictly more acceptable thans” := (Aat∧ ¬Aas)∨(¬Rat∧Ras)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Defining acceptance
B.Doxastic acceptance:
Aat:=�cϕ∈sub(t)Baϕ Rat:�cϕ∈sub(t)Ba¬ϕ
Fact: �Rat→ ¬Aatbut�¬Aat→Rat
A>a(t, s) :≈“a considerststrictly more acceptable thans” := (Aat∧ ¬Aas)∨(¬Rat∧Ras)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Defining acceptance
B.Doxastic acceptance:
Aat:=�cϕ∈sub(t)Baϕ Rat:�cϕ∈sub(t)Ba¬ϕ
Fact: �Rat→ ¬Aatbut�¬Aat→Rat
A>a(t, s) :≈“a considerststrictly more acceptable thans” := (Aat∧ ¬Aas)∨(¬Rat∧Ras)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Defining acceptance
B.Doxastic acceptance:
Aat:=�cϕ∈sub(t)Baϕ Rat:�cϕ∈sub(t)Ba¬ϕ
Fact: �Rat→ ¬Aatbut�¬Aat→Rat
A>a(t, s) :≈“a considerststrictly more acceptable thans” := (Aat∧ ¬Aas)∨(¬Rat∧Ras)
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Intuitive principles
B.Doxastic acceptance:
Ifρ1, and aconsiders the premises of tas strictly more
acceptable (A>a(t, s)), then she should strictly prefert
(Pa>(t, s, ϕ)).
Ifρ1, and aconsiders both set of premises equally acceptable
(A≈a(t, s)) she should go on her evaluation.
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Intuitive principles
B.Doxastic acceptance:
Ifρ1, and aconsiders the premises of tas strictly more acceptable (A>a(t, s)), then she should strictly prefert
(P>
a (t, s, ϕ)).
Ifρ1, and aconsiders both set of premises equally acceptable
(A≈a(t, s)) she should go on her evaluation.
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Intuitive principles
B.Doxastic acceptance:
Ifρ1, and aconsiders the premises of tas strictly more acceptable (A>a(t, s)), then she should strictly prefert
(P>
a (t, s, ϕ)).
Ifρ1, and aconsiders both set of premises equally acceptable
(A≈a(t, s)) she should go on her evaluation.
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Intuitive principles
C.Expert opinion:
Ifρ2, and b(the ϕ-expert of a(Dabϕ)) considers the premises of tstrictly more acceptable (A>b(t, s)), thena should strictly prefert (Pa>(t, s, ϕ)).
Ifρ2, and b(the ϕ-expert of a(Dabϕ)) considers the premises of tequally acceptable (A≈b (t, s)), thenashould go on her
evaluation.
Ifρ2, and aknows no expert on ϕ(�b∈A¬Dabϕ), thenashould
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Intuitive principles
C.Expert opinion:
Ifρ2, and b(the ϕ-expert of a(Dabϕ)) considers the premises of tstrictly more acceptable (A>b(t, s)), thenashould strictly prefert (P>
a (t, s, ϕ)).
Ifρ2, and b(the ϕ-expert of a(Dabϕ)) considers the premises of tequally acceptable (A≈b (t, s)), thenashould go on her
evaluation.
Ifρ2, and aknows no expert on ϕ(�b∈A¬Dabϕ), thenashould
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Intuitive principles
C.Expert opinion:
Ifρ2, and b(the ϕ-expert of a(Dabϕ)) considers the premises of tstrictly more acceptable (A>b(t, s)), thenashould strictly prefert (P>
a (t, s, ϕ)).
Ifρ2, and b(the ϕ-expert of a(Dabϕ)) considers the premises of tequally acceptable (A≈b (t, s)), thenashould go on her
evaluation.
Ifρ2, and aknows no expert on ϕ(�b∈A¬Dabϕ), thenashould
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Intuitive principles
C.Expert opinion:
Ifρ2, and b(the ϕ-expert of a(Dabϕ)) considers the premises of tstrictly more acceptable (A>b(t, s)), thenashould strictly prefert (P>
a (t, s, ϕ)).
Ifρ2, and b(the ϕ-expert of a(Dabϕ)) considers the premises of tequally acceptable (A≈b (t, s)), thenashould go on her
evaluation.
Ifρ2, and aknows no expert on ϕ(�b∈A¬Dabϕ), thenashould
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Intuitive principles
C.Expert opinion:
ρ3 :=ρ1∧((Dabϕ∧A≈b (t, s))∨(
�
b∈A
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Intuitive principles
C.Expert opinion:
ρ3 :=ρ1∧((Dabϕ∧A≈b (t, s))∨(
�
b∈A
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Intuitive principles
C.Expert opinion:
ρ3 :=ρ1∧((Dabϕ∧A≈b (t, s))∨(
�
b∈A
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Intuitive principles
D.Simplicity of arguments:
Ifρ3, and t is simpler thans(Com<(t, s)), thenashould prefer ttos(Pa>(t, s, ϕ).
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Intuitive principles
D.Simplicity of arguments:
Ifρ3, and t is simpler thans(Com<(t, s)), thenashould prefer
ttos(Pa>(t, s, ϕ).
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Intuitive principles
D.Simplicity of arguments:
Ifρ3, and t is simpler thans(Com<(t, s)), thenashould prefer
ttos(Pa>(t, s, ϕ).
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Intuitive principles
D.Simplicity of arguments:
Ifρ3, and t is simpler thans(Com<(t, s)), thenashould prefer
ttos(Pa>(t, s, ϕ).
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
P
a≥(
t, s, ϕ
) operator
The following formula allows us to capture all the principles above:
Pa≥(t, s, ϕ) := (¬(t�ϕ)∧ ¬(s�ϕ))∨(t�ϕ∧ ¬(s�ϕ))
∨(ρ1∧A>a(t, s))
∨(ρ2∧Dabϕ∧A>b (t, s))
∨(ρ3∧Com≤(t, s))
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
P
a≥(
t, s, ϕ
) operator
The following formula allows us to capture all the principles above:
Pa≥(t, s, ϕ) := (¬(t�ϕ)∧ ¬(s�ϕ))∨(t�ϕ∧ ¬(s�ϕ))
∨(ρ1∧A>a(t, s))
∨(ρ2∧Dabϕ∧A>b (t, s))
∨(ρ3∧Com≤(t, s))
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
P
a≥(
t, s, ϕ
) operator
The following formula allows us to capture all the principles above:
Pa≥(t, s, ϕ) := (¬(t�ϕ)∧ ¬(s�ϕ))∨(t�ϕ∧ ¬(s�ϕ))
∨(ρ1∧A>a(t, s))
∨(ρ2∧Dabϕ∧A>b (t, s))
∨(ρ3∧Com≤(t, s))
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
P
a≥(
t, s, ϕ
) operator
The following formula allows us to capture all the principles above:
Pa≥(t, s, ϕ) := (¬(t�ϕ)∧ ¬(s�ϕ))∨(t�ϕ∧ ¬(s�ϕ))
∨(ρ1∧A>a(t, s))
∨(ρ2∧Dabϕ∧A>b (t, s))
∨(ρ3∧Com≤(t, s))
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Justifying the introduction of
E
at
Preferential omniscience
�Pa≥(cϕ, t, ϕ) for any t∈ T
Intuitive reading: mono-premise arguments asserting the conclusion are always better.
Cause: definition of Pa≥ + standard omniscience
Solution: Pae≥(t, s, ϕ) :=Pa≥(t, s, ϕ)∧Eat∧Eas
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Justifying the introduction of
E
at
Preferential omniscience
�Pa≥(cϕ, t, ϕ) for any t∈ T
Intuitive reading: mono-premise arguments asserting the conclusion are always better.
Cause: definition of Pa≥ + standard omniscience
Solution: Pae≥(t, s, ϕ) :=Pa≥(t, s, ϕ)∧Eat∧Eas
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Justifying the introduction of
E
at
Preferential omniscience
�Pa≥(cϕ, t, ϕ) for any t∈ T
Intuitive reading: mono-premise arguments asserting the conclusion are always better.
Cause: definition of Pa≥ + standard omniscience
Solution: Pae≥(t, s, ϕ) :=Pa≥(t, s, ϕ)∧Eat∧Eas
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Justifying the introduction of
E
at
Preferential omniscience
�Pa≥(cϕ, t, ϕ) for any t∈ T
Intuitive reading: mono-premise arguments asserting the conclusion are always better.
Cause: definition of Pa≥ + standard omniscience
Solution: Pae≥(t, s, ϕ) :=Pa≥(t, s, ϕ)∧Eat∧Eas
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Justifying the introduction of
E
at
Preferential omniscience
�Pa≥(cϕ, t, ϕ) for any t∈ T
Intuitive reading: mono-premise arguments asserting the conclusion are always better.
Cause: definition of Pa≥ + standard omniscience
Solution: Pae≥(t, s, ϕ) :=Pa≥(t, s, ϕ)∧Eat∧Eas
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Justifying the introduction of
E
at
Preferential omniscience
�Pa≥(cϕ, t, ϕ) for any t∈ T
Intuitive reading: mono-premise arguments asserting the conclusion are always better.
Cause: definition of Pa≥ + standard omniscience
Solution: Pae≥(t, s, ϕ) :=Pa≥(t, s, ϕ)∧Eat∧Eas
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
Justifying the introduction of
E
at
Preferential omniscience
�Pa≥(cϕ, t, ϕ) for any t∈ T
Intuitive reading: mono-premise arguments asserting the conclusion are always better.
Cause: definition of Pa≥ + standard omniscience
Solution: Pae≥(t, s, ϕ) :=Pa≥(t, s, ϕ)∧Eat∧Eas
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
An example
� Anne is looking for her best argument to justify r:≈“it has rained”.
� She has no direct arguments cr available, since her office
has no windows (cr∈/Ea(w)).
� She considers twomodus ponens based arguments (· arguments) with the following premises “If streets are wet, the it has rained (cw→r). And it is the case that streets are
wet (cw)” and “If humidity is high, the it has rained
(ch→r). And it is the case that air humidity is high (ch)”.
Introduction The logicAE Modelling argument evaluation inAE Conclusions and future work References
An example
� Anne is looking for her best argument to justify r:≈“it has rained”.
� She has no direct arguments cr available, since her office
has no windows (cr∈/Ea(w)).
� She considers twomodus ponens based arguments (· arguments) with the following premises “If streets are wet, the it has rained (cw→r). And it is the case that streets are
wet (cw)” and “If humidity is high, the it has rained
(ch→r). And it is the case that air humidity is high (ch)”.