pith. sign in

arxiv: 1907.00352 · v1 · pith:PWGHHWNHnew · submitted 2019-06-30 · ⚛️ physics.soc-ph · cs.SI

Instability of social network dynamics with stubborn links

Somaye Sheykhali , Amir Hossein Darooneh , Gholam Reza Jafari This is my paper

Pith reviewed 2026-05-25 12:30 UTC · model grok-4.3

classification ⚛️ physics.soc-ph cs.SI
keywords signed networksstructural balancestubborn linkssocial dynamicsnetwork topologybalance evolutionantagonistic interactions
0
0 comments X

The pith

Locations of stubborn links shape signed network balance more than their count does.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper examines signed social networks where agents flip sentiments to reduce imbalance according to structural balance theory, but some links are stubborn and never change. It tests whether the fraction or the placement of these stubborn links controls whether the whole network reaches a balanced state. Both math and simulations show that placement matters far more: many stubborn links spread across the network still allow balance to emerge, while a small number clustered into five communities keeps the network unbalanced. This reveals that the topology created by the stubborn links overrides their raw quantity in setting the final global balance level.

Core claim

The paper claims that the global level of balance of the network is more influenced by the locations of stubborn links in the resulting network topology than by the fraction of stubborn links. Even with a large fraction of stubborn links the network would evolve towards a balanced state. On the other hand, if a small fraction of stubborn links are clustered in five stubborn communities, the network evolves to an unbalanced state.

What carries the argument

Stubborn links defined as extreme antagonistic interactions that resist all updates, embedded inside the iterative process of agents inverting sentiments to lower imbalance.

If this is right

  • Networks with stubborn links spread out reach a balanced state regardless of how many such links exist.
  • Clustering even a small number of stubborn links into exactly five communities produces a persistently unbalanced network.
  • The topology formed by stubborn links determines the balance outcome more strongly than their total count.
  • Poorly balanced final states consist of multiple antagonistic groups created by the clustered stubborn links.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Targeted placement of stubborn actors could be used to either stabilize or destabilize opinion dynamics in applied settings.
  • Real-world networks with stubborn users concentrated in a handful of tight groups may resist balance more than diffuse stubbornness would suggest.
  • Models could test whether the critical cluster count of five is special or whether other small integers produce similar instability.

Load-bearing premise

Agents always invert their sentiments to reduce imbalance unless a link is stubborn and therefore completely fixed and excluded from the update rule.

What would settle it

A simulation that places the same small fraction of stubborn links into four communities instead of five and checks whether the final balance level rises above the unbalanced threshold reported for five clusters.

Figures

Figures reproduced from arXiv: 1907.00352 by Amir Hossein Darooneh, Gholam Reza Jafari, Somaye Sheykhali.

Figure 1
Figure 1. Figure 1: Different balanced and unbalanced configurations of a triangle. Solid lines with ’+’ label represent positive [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Schematic pattern of a network with stubborn links (red dash lines) clustered in stubborn communities (blue [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: (A) Energy as a function of the fraction of stubborn links for a network with N = 100 nodes. Each point represents the final state of a single realization. The red points represent the energy reached by networks with a random displacement of stubborn links in the network. The blue points represent the extreme values of the energy of the final state when stubborn links form communities. The blue solid lines… view at source ↗
Figure 4
Figure 4. Figure 4: (A) Distribution of energy of the final state. (B) Distribution of number of triangles for any of the four types. For networks with N = 100 nodes and 1000 realizations. For energy reduction, the distribution of balanced and unbalanced triangles should be taken explicitly into account. In order to visually compare how the energy and the number of balanced and unbalanced triangles change as the fraction of s… view at source ↗
Figure 5
Figure 5. Figure 5: Numerical solution of energy (E) as a function of number of clusters (m), for a network of N=100 nodes. [PITH_FULL_IMAGE:figures/full_fig_p006_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: S-S distribution of stubborn degree for a network with [PITH_FULL_IMAGE:figures/full_fig_p007_6.png] view at source ↗
read the original abstract

This paper studies the signed networks in the presence of stubborn links, based on the structural balance theory. Each agent in the network has a mixture of positive and negative links represent friendly and antagonistic interactions and his stubbornness about interactions. Structural balance theory affirms that in signed social networks with simultaneous friendly/hostile interactions, there is a general tendency of evolving over time to reduce the tensions. From this perspective, individuals iteratively invert their own sentiments to reduce the felt tensions induced by imbalance. In this paper, we investigate the consequences of the agents' stubbornness on their interactions. We define stubbornness as an extreme antagonistic interaction which is resistant to change. In the current paper, we investigated if the presence of stubborn links renders an impact on the balance state of the network and whether or not the degree of balance in a signed network depends on the location of stubborn links. Our results show that a poorly balanced configuration consists of multiple antagonistic groups. Both analytical and simulation results demonstrate that the global level of balance of the network is more influenced by the locations of stubborn links in the resulting network topology than by the fraction of stubborn links. This means that even with a large fraction of stubborn links the network would evolve towards a balanced state. On the other hand, if a small fraction of stubborn links are clustered in five stubborn communities, the network evolves to an unbalanced state.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 2 minor

Summary. The manuscript studies signed networks under structural balance theory with the addition of stubborn links, defined as extreme antagonistic interactions that are fixed and do not participate in the update rule. Agents iteratively flip sentiments to reduce imbalance. Both analytical results and simulations are used to argue that the locations of stubborn links within the resulting topology influence the global balance level more strongly than the fraction of such links: large fractions can still permit evolution to a balanced state, whereas small fractions clustered into five stubborn communities drive the network to an unbalanced state.

Significance. If the central claim holds, the work demonstrates that topology and clustering of fixed antagonistic ties can override density in determining whether a signed network reaches structural balance. This adds a topological dimension to models of social dynamics with resistant agents and could be relevant for understanding persistent polarization. The combination of analysis and simulation is noted as a positive feature.

major comments (1)
  1. [Abstract] The central claim that location dominates fraction requires explicit quantitative support (e.g., plots or tables of the balance metric versus fraction for fixed versus clustered placements). The abstract states the result but does not indicate the precise balance measure, the number of realizations, or statistical controls used to establish dominance.
minor comments (2)
  1. Clarify the precise update rule for non-stubborn links and the mathematical definition of the global balance metric used in both the analysis and the simulations.
  2. Specify the network sizes, initial conditions, and convergence criteria employed in the simulations to allow reproducibility.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the positive assessment and constructive feedback. We address the major comment below.

read point-by-point responses

  1. Referee: [Abstract] The central claim that location dominates fraction requires explicit quantitative support (e.g., plots or tables of the balance metric versus fraction for fixed versus clustered placements). The abstract states the result but does not indicate the precise balance measure, the number of realizations, or statistical controls used to establish dominance.

    Authors: We agree that the abstract can be strengthened with more explicit quantitative details. The manuscript body (Sections 3-4) already contains the supporting analysis and simulations: the balance metric is the fraction of balanced triads, with results averaged over 100 realizations and shown with standard deviations for both clustered (five communities) and spread-out placements across fractions from 0.05 to 0.3. We will revise the abstract to specify this metric, the number of realizations, and the key comparisons demonstrating that clustered placement drives imbalance even at low fractions while spread-out placement permits balance at higher fractions. revision: yes

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper defines an explicit dynamical model on signed networks under structural balance, with agents iteratively inverting sentiments according to a tension-reduction rule and stubborn links implemented as fixed extreme antagonistic interactions that do not update. Both analytical arguments and direct simulations of this process are used to compare the effects of stubborn-link fraction versus their topological clustering. No fitted parameters are renamed as predictions, no self-citations supply load-bearing uniqueness theorems or ansatzes, and the balance metric is computed directly from the simulated or analyzed configurations rather than being defined in terms of itself. The derivation chain is therefore self-contained against the stated update rule and does not reduce outputs to inputs by construction.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Abstract-only review; full model specification unavailable. No free parameters, invented entities, or additional axioms beyond the stated domain assumption are identifiable.

axioms (1)
  • domain assumption Structural balance theory: signed networks evolve by agents inverting sentiments to reduce imbalance tensions.
    Explicitly invoked as the basis for the iterative dynamics and stubborn-link definition.

pith-pipeline@v0.9.0 · 5780 in / 1202 out tokens · 27787 ms · 2026-05-25T12:30:19.021422+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

46 extracted references · 46 canonical work pages

  1. [1]

    Attitudes and cognitive organization

    Fritz Heider. Attitudes and cognitive organization. The Journal of Psychology, 21(1):107–112, 1946

    work page 1946

  2. [2]

    Cambridge university press, Cambridge, 1994

    Stanley Wasserman and Katherine Faust.Social network analysis: Methods and applications, volume 8. Cambridge university press, Cambridge, 1994

    work page 1994

  3. [3]

    Networks, Crowds, and Markets: Reasoning about a Highly Connected World

    David Easley and John Kleinberg. Networks, Crowds, and Markets: Reasoning about a Highly Connected World . Cambridge University Press, Cambridge, 2010

    work page 2010

  4. [4]

    Computing global structural balance in large-scale signed social networks

    Giuseppe Facchetti, Giovanni Iacono, and Claudio Altafini. Computing global structural balance in large-scale signed social networks. Proceedings of the National Academy of Sciences , 108(52):20953–20958, 2011

    work page 2011

  5. [5]

    An evolutionary game approach for determination of the structural conflicts in signed networks

    Shaolin Tan and Jinhu Lü. An evolutionary game approach for determination of the structural conflicts in signed networks. Proceedings of the National Academy of Sciences , 6:22022, 2016

    work page 2016

  6. [6]

    Structural balance: a generalization of heider’s theory

    Dorwin Cartwright and Frank Harary. Structural balance: a generalization of heider’s theory. American Psycho- logical Association, 63(5):277, 1956. 7 A PREPRINT - J ULY 2, 2019

    work page 1956

  7. [7]

    Structural balance in signed networks: Separating the probability to interact from the tendency to fight

    Jürgen Lerner. Structural balance in signed networks: Separating the probability to interact from the tendency to fight. Social Networks, 45:66–77, 2016

    work page 2016

  8. [8]

    Statistical physics of social dynamics

    Claudio Castellano, Santo Fortunato, and Vittorio Loreto. Statistical physics of social dynamics. Reviews of modern physics, 81(2):591, 2009

    work page 2009

  9. [9]

    The Dynamics of Two-Party Politics: Party Structures and the Management of Competition

    Alan Ware. The Dynamics of Two-Party Politics: Party Structures and the Management of Competition . Oxford University Press„ 2009

    work page 2009

  10. [10]

    Saeedian, N

    M. Saeedian, N. Azimi-Tafreshi, G. R. Jafari, and J. Kertesz. Epidemic spreading on evolving signed networks. Phys. Rev. E, 95(2):022314, 2017

    work page 2017

  11. [11]

    Robust inference with multiway clustering

    A Colin Cameron, Jonah Gelbach B, and Douglas L Miller. Robust inference with multiway clustering. Journal of Business & Economic Statistics , 29(2):238–249, 2011

    work page 2011

  12. [12]

    Hugo Saiz, Jesús Gómez Gardeñes, Paloma Nuche, Andrea Girón, Yolanda Pueyo, and Concepción L. Alados. Clustering and structural balance in graphs. Ecography, 40(6):733–741, 2017

    work page 2017

  13. [13]

    Structural balance in the social networks of a wild mammal

    Amiyaal Ilany, Adi Barocas, Lee Koren, Michael Kam, and Eli Geffen. Structural balance in the social networks of a wild mammal. Animal Behaviour, 85(6):1397–1405, 2013

    work page 2013

  14. [14]

    The structural balance analysis of complex dynamical networks based on nodes’ dynamical couplings

    Zilin Gao and Yinhe Wang. The structural balance analysis of complex dynamical networks based on nodes’ dynamical couplings. PloS one, 13(1):e0191941, 2018

    work page 2018

  15. [15]

    Measuring partial balance in signed networks

    Samin Aref and Mark C Wilson. Measuring partial balance in signed networks. Journal of Complex Networks, 6(4):566–595, 2017

    work page 2017

  16. [16]

    Cantwell, and M

    Alec Kirkley, George T. Cantwell, and M. E. J. Newman. Balance in signed networks. Phys. Rev. E, 99(1):012320, 2019

    work page 2019

  17. [17]

    Facchetti, G

    G. Facchetti, G. Iacono C., and Altafini. Computing global structural balance in large-scale signed social networks. Proceedings of the National Academy of Sciences , 108(52):120953–20958, 2011

    work page 2011

  18. [18]

    Statistical physics of balance theory

    Andres M Belaza, Kevin Hoefman, Jan Ryckebusch, Aaron Bramson, Milan van den Heuvel, and Koen Schoors. Statistical physics of balance theory. PLoS one, 12(8):e0183696, 2017

    work page 2017

  19. [19]

    Krapivsky, and Sidney Redner

    Tibor Antal, Pavel L. Krapivsky, and Sidney Redner. Dynamics of social balance on networks. Physical Review E, 72(3):036121, 2005

    work page 2005

  20. [20]

    Dynamics of opinion forming in structurally balanced social networks

    Claudio Altafini. Dynamics of opinion forming in structurally balanced social networks. PLoS ONE, 7(6):e38135, 2012

    work page 2012

  21. [21]

    Optimizing transformations of structural balance in signed networks with potential relationships

    Haifeng Du, Xiaochen He, Shanfeng Wang, Maoguo Gong, and Marcus W Feldman. Optimizing transformations of structural balance in signed networks with potential relationships. Physica A: Statistical Mechanics and its Applications, 465:414–424, 2017

    work page 2017

  22. [22]

    The heider balance: A continuous approach

    Krzysztof Kułakowski, Przemysław Gawro´nski, and Piotr Gronek. The heider balance: A continuous approach. International Journal of Modern Physics C , 16(5):707–716, 2015

    work page 2015

  23. [23]

    Hedayatifar, F

    L. Hedayatifar, F. Hassanibesheli, A. Shirazi, S. V . Farahani, and G. R. Jafari. Pseudo paths towards minimum energy states in network dynamics. Physica A, 483(7307):109–116, 2017

    work page 2017

  24. [24]

    Reversing structural balance in signed networks

    Haifeng Dua, Xiaochen He, Jingjing Wanga, and Marcus W.Feldman. Reversing structural balance in signed networks. Physica A: Statistical Mechanics and its Applications , 503:780–792, 2018

    work page 2018

  25. [25]

    Gain and loss of esteem, direct reciprocity and heider balance

    Forough Hassanibesheli, Leila Hedayatifar, Przemysław Gawro´nski, Maria Stojkow, Dorota ˙Zuchowska Skiba, and Krzysztof Kułakowski. Gain and loss of esteem, direct reciprocity and heider balance. Physica A: Statistical Mechanics and its Applications , 468:0378–4371, 2017

    work page 2017

  26. [26]

    Social stability and extended social balance—quantifying the role of inactive links in social networks

    Andres M Belaza, Jan Ryckebusch, Aaron Bramson, Corneel Casert, Kevin Hoefman, Koen Schoors, Milanvan den Heuvel, and Benjamin Vandermarliere. Social stability and extended social balance—quantifying the role of inactive links in social networks. Physica A: Statistical Mechanics and its Applications , 518:270–284, 2019

    work page 2019

  27. [27]

    Debenedetti and Frank H

    Pablo G. Debenedetti and Frank H. Stillinger. Supercooled liquids and the glass transition. Nature, 410(6825):259, 2001

    work page 2001

  28. [28]

    nergy landscape and dynamics of brain activity during human bistable perception

    Takamitsu Watanabe, Naoki Masuda, Fukuda Megumi, Ryota Kanai, and Geraint Rees. nergy landscape and dynamics of brain activity during human bistable perception. Nature communications, 5(6):4765, 2014

    work page 2014

  29. [29]

    Marvel, Steven H

    Seth A. Marvel, Steven H. Strogatz, and Jon M. Kleinberg. Energy landscape of social balance. Frontiers in Physics, 103(19):198701, 2009

    work page 2009

  30. [30]

    Opinion dynamics in social networks with stubborn agents: Equilibrium and convergence rate

    Javad Ghaderi and Rayadurgam Srikant. Opinion dynamics in social networks with stubborn agents: Equilibrium and convergence rate. Automatica, 50(12):3209–3215, 2014. 8 A PREPRINT - J ULY 2, 2019

    work page 2014

  31. [31]

    Binary opinion dynamics with stubborn agents

    Ercan Yildiz, Asuman Ozdaglar, Daron Acemoglu, Amin Saberi, and Anna Scaglione. Binary opinion dynamics with stubborn agents. ACM Transactions on Economics and Computation (TEAC) , 1(4):19, 2013

    work page 2013

  32. [32]

    Spectral properties of the grounded laplacian matrix with applications to consensus in the presence of stubborn agents

    Mohammad Pirani and Shreyas Sundaram. Spectral properties of the grounded laplacian matrix with applications to consensus in the presence of stubborn agents. In The title of the conference proceedings , pages 2160–2165. American Control Conference, 2014

    work page 2014

  33. [33]

    Slow-down or speed-up of inter-and intra-cluster diffusion of controversial knowledge in stubborn communities based on a small world network

    Marcel Ausloos. Slow-down or speed-up of inter-and intra-cluster diffusion of controversial knowledge in stubborn communities based on a small world network. Frontiers in Physics, 3:43, 2015

    work page 2015

  34. [34]

    Friedkin, and Francesco Bullo

    Anahita Mirtabatabaei, Peng Jia, Noah E. Friedkin, and Francesco Bullo. On the reflected appraisals dynamics of influence networks with stubborn agents. In The title of the conference proceedings , pages 3978–3983. American Control Conference, IEEE, 2014

    work page 2014

  35. [35]

    Signed networks in social media

    Jure Leskovec, Daniel Huttenlocher, and Jon Kleinberg. Signed networks in social media. In Proceedings of the SIGCHI conference on human factors in computing systems , pages 1361–1370. ACM, 2010

    work page 2010

  36. [36]

    The micro-macro link for the theory of structural balance

    Arnout Van de Rijt. The micro-macro link for the theory of structural balance. The Journal of Mathematical Sociology, 35(1-3):94–113, 2011

    work page 2011

  37. [37]

    Walk-based measure of balance in signed networks: Detecting lack of balance in social networks

    Ernesto Estrada and Michele Benzi. Walk-based measure of balance in signed networks: Detecting lack of balance in social networks. Physical Review E, 90(4):042802, 2014

    work page 2014

  38. [38]

    A partitioning approach to structural balance

    Patrick Doreian and Andrej Mrvar. A partitioning approach to structural balance. Social networks, 18(2):149–168, 1996

    work page 1996

  39. [39]

    Bagrow, and Sune Lehmann

    Yong-Yeol Ahn, James P. Bagrow, and Sune Lehmann. Link communities reveal multiscale complexity in networks. Nature, 466(7307):761, 2010

    work page 2010

  40. [40]

    Multirelational organization of large-scale social networks in an online world

    Michael Szell, Renaud Lambiotte, and Stefan Thurner. Multirelational organization of large-scale social networks in an online world. Proceedings of the National Academy of Sciences , 107(31):13636–13641, 2010

    work page 2010

  41. [41]

    James A. Davis. Clustering and structural balance in graphs. Human Relations, 20(2):181–187, 1967

    work page 1967

  42. [42]

    Community structure of complex networks

    Hua-Wei Shen. Community structure of complex networks . Springer Science & Business Media, 2013

    work page 2013

  43. [43]

    Mark E. J. Newman, Duncan J. Watts, and Steven HStrogatz. Random graph models of social networks. Proceedings of the National Academy of Sciences , 99(1):2566–2572, 2002

    work page 2002

  44. [44]

    Glassy states of aging social networks

    Foroogh Hassanibesheli, Leila Hedayatifar, Hadise Safdari, Marcel Ausloos, and Gholamreza Jafari. Glassy states of aging social networks. Entropy, 19(6):246, 2017

    work page 2017

  45. [45]

    Random graphs

    Béla Bollobás. Random graphs. Number 73. Cambridge university press, 2001

    work page 2001

  46. [46]

    Ausloos, and Araceli Noemi Proto

    Leonidas Facundo Caram, Cesar Federico Caiafa, M. Ausloos, and Araceli Noemi Proto. Cooperative peer-to-peer multiagent-based systems. Physical Review E - Statistical, Nonlinear , and Soft Matter Physics , 92(2):022805, 2015. 9

    work page 2015