Semantic networks are used in natural language processing applications such as semantic parsing and word-sense disambiguation. Semantic networks can also be used as a method to analyze large texts
and identify the main themes and topics (e.g., of social media posts),
to reveal biases (e.g., in news coverage), or even to map an entire
research field.
History
Examples of the use of semantic networks in logic, directed acyclic
graphs as a mnemonic tool, dates back centuries. The earliest documented
use being the Greek philosopher Porphyry's commentary on Aristotle's
categories in the third century AD.
In computing history, "Semantic Nets" for the propositional calculus were first implemented for computers by Richard H. Richens of the Cambridge Language Research Unit in 1956 as an "interlingua" for machine translation of natural languages. Although the importance of this work and the CLRU was only belatedly realized.
Semantic networks were also independently implemented by Robert F. Simmons and Sheldon Klein, using the first order predicate calculus as a base,
after being inspired by a demonstration of Victor Yngve. The "line of
research was originated by the first President of the Association
[Association for Computational Linguistics], Victor Yngve, who in 1960
had published descriptions of algorithms for using a phrase structure
grammar to generate syntactically well-formed nonsense sentences.
Sheldon Klein and I about 1962-1964 were fascinated by the technique and
generalized it to a method for controlling the sense of what was
generated by respecting the semantic dependencies of words as they
occurred in text." Other researchers, most notably M. Ross Quillian and others at System Development Corporation
helped contribute to their work in the early 1960s as part of the
SYNTHEX project. It's from these publications at SDC that most modern
derivatives of the term "semantic network" cite as their background.
Later prominent works were done by Allan M. Collins and Quillian (e.g., Collins and Quillian; Collins and Loftus Quillian. Still later in 2006, Hermann Helbig fully described MultiNet.
In the late 1980s, two Netherlands universities, Groningen and Twente, jointly began a project called Knowledge Graphs,
which are semantic networks but with the added constraint that edges
are restricted to be from a limited set of possible relations, to
facilitate algebras on the graph. In the subsequent decades, the distinction between semantic networks and knowledge graphs was blurred. In 2012, Google gave their knowledge graph the name Knowledge Graph.
The Semantic Link Network was systematically studied as a social
semantics networking method. Its basic model consists of semantic nodes,
semantic links between nodes, and a semantic space that defines the
semantics of nodes and links and reasoning rules on semantic links. The
systematic theory and model was published in 2004. This research direction can trace to the definition of inheritance rules for efficient model retrieval in 1998 and the Active Document Framework ADF. Since 2003, research has developed toward social semantic networking. This work is a systematic innovation at the age of the World Wide Web
and global social networking rather than an application or simple
extension of the Semantic Net (Network). Its purpose and scope are
different from that of the Semantic Net (or network). The rules for reasoning and evolution and automatic discovery of
implicit links play an important role in the Semantic Link Network. Recently it has been developed to support Cyber-Physical-Social Intelligence. It was used for creating a general summarization method. The self-organised Semantic Link Network was integrated with a
multi-dimensional category space to form a semantic space to support
advanced applications with multi-dimensional abstractions and
self-organised semantic links It has been verified that Semantic Link Network play an important role
in understanding and representation through text summarisation
applications. Semantic Link Network has been extended from cyberspace to
cyber-physical-social space. Competition relation and symbiosis relation
as well as their roles in evolving society were studied in the emerging
topic: Cyber-Physical-Social Intelligence
More specialized forms of semantic networks has been created for
specific use. For example, in 2008, Fawsy Bendeck's PhD thesis
formalized the Semantic Similarity Network (SSN) that contains specialized relationships and propagation algorithms to simplify the semantic similarity representation and calculations.
Basics of semantic networks
A semantic network is used when one has knowledge that is best understood as a set of concepts that are related to one another.
Most semantic networks are cognitively based. They also consist
of arcs and nodes which can be organized into a taxonomic hierarchy.
Semantic networks contributed ideas of spreading activation, inheritance, and nodes as proto-objects.
An example of a semantic network is WordNet, a lexical database of English. It groups English words into sets of synonyms called synsets,
provides short, general definitions, and records the various semantic
relations between these synonym sets. Some of the most common semantic
relations defined are meronymy (A is a meronym of B if A is part of B), holonymy (B is a holonym of A if B contains A), hyponymy (or troponymy) (A is subordinate of B; A is kind of B), hypernymy (A is superordinate of B), synonymy (A denotes the same as B) and antonymy (A denotes the opposite of B).
It is also possible to represent logical descriptions using semantic networks such as the existential graphs of Charles Sanders Peirce or the related conceptual graphs of John F. Sowa. These have expressive power equal to or exceeding standard first-order predicate logic.
Unlike WordNet or other lexical or browsing networks, semantic
networks using these representations can be used for reliable automated
logical deduction. Some automated reasoners exploit the graph-theoretic
features of the networks during processing.
Other examples of semantic networks are Gellish models. Gellish English with its Gellish English dictionary, is a formal language
that is defined as a network of relations between concepts and names of
concepts. Gellish English is a formal subset of natural English, just
as Gellish Dutch is a formal subset of Dutch, whereas multiple languages
share the same concepts. Other Gellish networks consist of knowledge
models and information models that are expressed in the Gellish
language. A Gellish network is a network of (binary) relations between
things. Each relation in the network is an expression of a fact that is
classified by a relation type. Each relation type itself is a concept
that is defined in the Gellish language dictionary. Each related thing
is either a concept or an individual thing that is classified by a
concept. The definitions of concepts are created in the form of
definition models (definition networks) that together form a Gellish
Dictionary. A Gellish network can be documented in a Gellish database
and is computer interpretable.
SciCrunch
is a collaboratively edited knowledge base for scientific resources. It
provides unambiguous identifiers (Research Resource IDentifiers or
RRIDs) for software, lab tools etc. and it also provides options to
create links between RRIDs and from communities.
Another example of semantic networks, based on category theory, is ologs. Here each type is an object, representing a set of things, and each arrow is a morphism, representing a function. Commutative diagrams also are prescribed to constrain the semantics.
In the social sciences people sometimes use the term semantic network to refer to co-occurrence networks.
Software tools
There are also elaborate types of semantic networks connected with corresponding sets of software tools used for lexicalknowledge engineering, like the Semantic Network Processing System (SNePS) of Stuart C. Shapiro or the MultiNet paradigm of Hermann Helbig, especially suited for the semantic representation of natural language expressions and used in several NLP applications.
Semantic networks are used in specialized information retrieval tasks, such as plagiarism detection. They provide information on hierarchical relations in order to employ semantic compression to reduce language diversity and enable the system to match word meanings, independently from sets of words used.
The Knowledge Graph proposed by Google in 2012 is actually an application of semantic network in search engine.
Modeling multi-relational data like semantic networks in low-dimensional spaces through forms of embedding
has benefits in expressing entity relationships as well as extracting
relations from mediums like text. There are many approaches to learning
these embeddings, notably using Bayesian clustering frameworks or
energy-based frameworks, and more recently, TransE (NIPS 2013). Applications of embedding knowledge base data include Social network analysis and Relationship extraction.
Systems biology is the computational and mathematical analysis and modeling of complex biological systems. It is a biology-based
interdisciplinary field of study that focuses on complex interactions
within biological systems, using a holistic approach (holism instead of the more traditional reductionism) to biological research. This multifaceted research domain necessitates the collaborative
efforts of chemists, biologists, mathematicians, physicists, and
engineers to decipher the biology of intricate living systems by merging
various quantitative molecular measurements with carefully constructed
mathematical models. It represents a comprehensive method for
comprehending the complex relationships within biological systems. In
contrast to conventional biological studies that typically center on
isolated elements, systems biology seeks to combine different biological
data to create models that illustrate and elucidate the dynamic
interactions within a system. This methodology is essential for
understanding the complex networks of genes, proteins, and metabolites that influence cellular activities and the traits of organisms. One of the aims of systems biology is to model and discover emergent
properties, of cells, tissues and organisms functioning as a system
whose theoretical description is only possible using techniques of
systems biology. By exploring how function emerges from dynamic interactions, systems
biology bridges the gaps that exist between molecules and physiological
processes.
As a paradigm, systems biology is usually defined in antithesis to the so-called reductionist paradigm (biological organisation), although it is consistent with the scientific method. The distinction between the two paradigms is referred to in these quotations: "the reductionist
approach has successfully identified most of the components and many of
the interactions but, unfortunately, offers no convincing concepts or
methods to understand how system properties emerge ... the pluralism of
causes and effects in biological networks is better addressed by
observing, through quantitative measures, multiple components
simultaneously and by rigorous data integration with mathematical
models." (Sauer et al.) "Systems biology ... is about putting together rather than taking
apart, integration rather than reduction. It requires that we develop
ways of thinking about integration that are as rigorous as our
reductionist programmes, but different. ... It means changing our
philosophy, in the full sense of the term." (Denis Noble)
As a series of operational protocols used for performing research, namely a cycle composed of theory, analytic or computational modelling
to propose specific testable hypotheses about a biological system,
experimental validation, and then using the newly acquired quantitative
description of cells or cell processes to refine the computational model
or theory. Since the objective is a model of the interactions in a system, the
experimental techniques that most suit systems biology are those that
are system-wide and attempt to be as complete as possible. Therefore, multiomics (transcriptomics, metabolomics, proteomics, etc) and high-throughput techniques are used to collect quantitative data for the construction and validation of models.
A comprehensive systems biology approach necessitates: (i) a
thorough characterization of an organism concerning its molecular
components, the interactions among these molecules, and how these
interactions contribute to cellular functions; (ii) a detailed
spatio-temporal molecular characterization of a cell (for example,
component dynamics, compartmentalization, and vesicle transport); and
(iii) an extensive systems analysis of the cell's 'molecular response'
to both external and internal perturbations. Furthermore, the data from
(i) and (ii) should be synthesized into mathematical models to test
knowledge by generating predictions (hypotheses), uncovering new
biological mechanisms, assessing the system's behavior derived from
(iii), and ultimately formulating rational strategies for controlling
and manipulating cells. To tackle these challenges, systems biology must
incorporate methods and approaches from various disciplines that have
not traditionally interfaced with one another. The emergence of multi-omics technologies has transformed systems
biology by providing extensive datasets that cover different biological
layers, including genomics, transcriptomics, proteomics, and
metabolomics. These technologies enable the large-scale measurement of
biomolecules, leading to a more profound comprehension of biological
processes and interactions. Increasingly, methods such as network analysis, machine learning, and
pathway enrichment are utilized to integrate and interpret multi-omics
data, thereby improving our understanding of biological functions and
disease mechanisms.
History
Shows
trends in systems biology research by presenting the number of articles
out of the top 30 cited systems biology papers during that time which
include a specific topic
Holism vs. Reductionism
It is challenging to trace the origins and beginnings of systems
biology. A comprehensive perspective on the human body was central to
the medical practices of Greek, Roman, and East Asian traditions, where
physicians and thinkers like Hippocrates believed that health and
illness were linked to the equilibrium or disruption of bodily fluids
known as humors. This holistic perspective persisted in the Western
world throughout the 19th and 20th centuries, with prominent
physiologists viewing the body as controlled by various systems,
including the nervous system, the gastrointestinal system, and the
cardiovascular system. In the latter half of the 20th century, however,
this way of thinking was largely supplanted by reductionism: To grasp how the body functions properly, one needed to comprehend the
role of each component, from tissues and cells to the complete set of
intracellular molecular building blocks.
In the 17th century, the triumphs of physics and the advancement
of mechanical clockwork prompted a reductionist viewpoint in biology,
interpreting organisms as intricate machines made up of simpler
elements.
Jan Smuts
(1870–1950), naturalist/philosopher and twice Prime Minister of South
Africa, coined the commonly used term holism. Whole systems such as
cells, tissues, organisms, and populations were proposed to have unique
(emergent) properties. It was impossible to try and reassemble the
behavior of the whole from the properties of the individual components,
and new technologies were necessary to define and understand the
behavior of systems.
Even though reductionism and holism are often contrasted with one
another, they can be synthesized. One must understand how organisms are
built (reductionism), while it is just as important to understand why
they are so arranged (systems; holism). Each provides useful insights
and answers different questions. However, the study of biological
systems requires knowledge about control and design paradigms, as well
as principles of structural stability, resilience, and robustness that
are not directly inferred from mechanistic information. More profound
insight will be gained by employing computer modeling to overcome the
complexity in biological systems.
Nevertheless, this perspective was consistently balanced by
thinkers who underscored the significance of organization and emergent
traits in living systems. This reductionist perspective has achieved
remarkable success, and our understanding of biological processes has
expanded with incredible speed and intensity. However, alongside these
extraordinary advancements, science gradually came to understand that
possessing complete information about molecular components alone would
not suffice to elucidate the workings of life: the individual components
rarely illustrate the function of a complex system. It is now commonly
recognized that we need approaches for reconstructing integrated systems
from their constituent parts and processes if we are to comprehend
biological phenomena and manipulate them in a thoughtful, focused way.
Origin of systems biology as a field
In 1968, the term "systems biology" was first introduced at a conference. Those within the discipline soon recognized—and this understanding
gradually became known to the wider public—that computational approaches
were necessary to fully articulate the concepts and potential of
systems biology. Specifically, these techniques needed to view
biological phenomena as complex, multi-layered, adaptive, and dynamic
systems. They had to account for transformations and intricate
nonlinearities, thereby allowing for the smooth integration of smaller
models ("modules") into larger, well-organized assemblies of models
within complex settings. It became clear that mathematics and
computation were vital for these methods. An acceleration of systems understanding came with the publication of
the first ground-breaking text compiling molecular, physiological, and
anatomical individuality in animals, which has been described as a revolution.
Initially, the wider scientific community was reluctant to accept
the integration of computational methods and control theory in the
exploration of living systems, believing that "biology was too complex
to apply mathematics." However, as the new millennium neared, this
viewpoint underwent a significant and lasting transformation. More scientists started working on integration of mathematical concepts
to understand and solve biological problems. Now, systems biology has
been widely applied in several fields including agriculture and medicine.
Approaches to systems biology
Top-down approach
Top-down systems biology identifies molecular interaction networks by
analyzing the correlated behaviors observed in large-scale 'omics'
studies. With the advent of 'omics', this top-down strategy has become a
leading approach. It begins with an overarching perspective of the
system's behavior – examining everything at once – by gathering
genome-wide experimental data and seeks to unveil and understand
biological mechanisms at a more granular level – specifically, the
individual components and their interactions. In this framework of
'top-down' systems biology, the primary goal is to uncover novel
molecular mechanisms through a cyclical process that initiates with
experimental data, transitions into data analysis and integration to
identify correlations among molecule concentrations and concludes with
the development of hypotheses regarding the co- and inter-regulation of
molecular groups. These hypotheses then generate new predictions of
correlations, which can be explored in subsequent experiments or through
additional biochemical investigations. The notable advantages of top-down systems biology lie in its potential
to provide comprehensive (i.e., genome-wide) insights and its focus on
the metabolome, fluxome, transcriptome, and/or proteome.
Top-down methods prioritize overall system states as influencing
factors in models and the computational (or optimality) principles that
govern the dynamics of the global system. For instance, while the
dynamics of motor control (neuro) emerge from the interactions of
millions of neurons, one can also characterize the neural motor system
as a sort of feedback control system, which directs a 'plant' (the body)
and guides movement by minimizing 'cost functions' (e.g., achieving
trajectories with minimal jerk).
Bottom-up approach
Bottom-up systems biology infers the functional characteristics that
may arise from a subsystem characterized with a high degree of
mechanistic detail using molecular techniques. This approach begins with
the foundational elements by developing the interactive behavior (rate
equation) of each component process (e.g., enzymatic processes) within a
manageable portion of the system. It examines the mechanisms through
which functional properties arise in the interactions of known
components. Subsequently, these formulations are combined to understand
the behavior of the system. The primary goal of this method is to
integrate the pathway models into a comprehensive model representing the
entire system - the top or whole. As research and understanding
advance, these models are often expanded by incorporating additional
processes with high mechanistic detail.
The bottom-up approach facilitates the integration and
translation of drug-specific in vitro findings to the in vivo human
context. This encompasses data collected during the early phases of drug
development, such as safety evaluations. When assessing cardiac safety,
a purely bottom-up modeling and simulation method entails
reconstructing the processes that determine exposure, which includes the
plasma (or heart tissue) concentration-time profiles and their
electrophysiological implications, ideally incorporating hemodynamic
effects and changes in contractility. Achieving this necessitates
various models, ranging from single-cell to advanced three-dimensional
(3D) multiphase models. Information from multiple in vitro systems that
serve as stand-ins for the in vivo absorption, distribution, metabolism,
and excretion (ADME) processes enables predictions of drug exposure,
while in vitro data on drug-ion channel interactions support the
translation of exposure to body surface potentials and the calculation
of important electrophysiological endpoints. The separation of data
related to the drug, system, and trial design, which is characteristic
of the bottom-up approach, allows for predictions of exposure-response
relationships considering both inter- and intra-individual variability,
making it a valuable tool for evaluating drug effects at a population
level. Numerous successful instances of applying physiologically based
pharmacokinetic (PBPK) modeling in drug discovery and development have
been documented in the literature.
According to the interpretation of systems biology as using large
data sets using interdisciplinary tools, a typical application is metabolomics, which is the complete set of all the metabolic products, metabolites, in the system at the organism, cell, or tissue level.
The molecular interactions within the cell are also studied, this is called interactomics. A discipline in this field of study is protein–protein interactions, although interactomics includes the interactions of other molecules. Neuroelectrodynamics, where the computer's or a brain's computing function as a dynamic system is studied along with its (bio)physical mechanisms; and fluxomics, measurements of the rates of metabolic reactions in a biological system (cell, tissue, or organism).
In approaching a systems biology problem there are two main
approaches. These are the top down and bottom up approach. The top down
approach takes as much of the system into account as possible and relies
largely on experimental results. The RNA-Seq
technique is an example of an experimental top down approach.
Conversely, the bottom up approach is used to create detailed models
while also incorporating experimental data. An example of the bottom up
approach is the use of circuit models to describe a simple gene network.
Various technologies utilized to capture dynamic changes in mRNA, proteins, and post-translational modifications. Mechanobiology, forces and physical properties at all scales, their interplay with other regulatory mechanisms; biosemiotics, analysis of the system of sign relations of an organism or other biosystems; Physiomics, a systematic study of physiome in biology.
Cancer systems biology is an example of the systems biology approach, which can be distinguished by the specific object of study (tumorigenesis and treatment of cancer). It works with the specific data (patient samples, high-throughput data with particular attention to characterizing cancer genome in patient tumour samples) and tools (immortalized cancer cell lines, mouse models of tumorigenesis, xenograft models, high-throughput sequencing methods, siRNA-based gene knocking down high-throughput screenings, computational modeling of the consequences of somatic mutations and genome instability). The long-term objective of the systems biology of cancer is ability to
better diagnose cancer, classify it and better predict the outcome of a
suggested treatment, which is a basis for personalized cancer medicine and virtual cancer patient
in more distant prospective. Significant efforts in computational
systems biology of cancer have been made in creating realistic
multi-scale in silico models of various tumours.
The systems biology approach often involves the development of mechanistic models, such as the reconstruction of dynamic systems from the quantitative properties of their elementary building blocks. For instance, a cellular network can be modelled mathematically using methods coming from chemical kinetics and control theory.
Due to the large number of parameters, variables and constraints in
cellular networks, numerical and computational techniques are often used
(e.g., flux balance analysis).
Other aspects of computer science, informatics, and statistics are also used in systems biology. These include new forms of computational models, such as the use of process calculi to model biological processes (notable approaches include stochastic π-calculus, BioAmbients, Beta Binders, BioPEPA, and Brane calculus) and constraint-based modeling; integration of information from the literature, using techniques of information extraction and text mining; development of online databases and repositories for sharing data and
models, approaches to database integration and software interoperability
via loose coupling
of software, websites and databases, or commercial suits; network-based
approaches for analyzing high dimensional genomic data sets. For
example, weighted correlation network analysis
is often used for identifying clusters (referred to as modules),
modeling the relationship between clusters, calculating fuzzy measures
of cluster (module) membership, identifying intramodular hubs, and for
studying cluster preservation in other data sets; pathway-based methods
for omics data analysis, e.g. approaches to identify and score pathways
with differential activity of their gene, protein, or metabolite
members. Much of the analysis of genomic data sets also include identifying
correlations. Additionally, as much of the information comes from
different fields, the development of syntactically and semantically
sound ways of representing biological models is needed.
Model and its types
Definition
A model serves as a conceptual depiction of objects or processes,
highlighting certain characteristics of these items or activities. A
model captures only certain facets of reality; however, when created
correctly, this limited scope is adequate because the primary goal of
modeling is to address specific inquiries. The saying, "essentially, all models are wrong, but some are useful,"
attributed to the statistician George Box, is a suitable principle for
constructing models.
Types of models
Boolean Models: These models are also known as logical models
and represent biological systems using binary states, allowing for the
analysis of gene regulatory networks and signaling pathways. They are
advantageous for their simplicity and ability to capture qualitative
behaviors.This
figure deals with the tool BooleSim which is used for simulating and
manipulating Boolean models. The given figure deals with a simple
synthetic repressilator (A&C) and the concerned output (time series)
is obtained using the tool BooleSim (B & D). The boxes represent
nodes and the arrow shows the relationship between them. Pointed and
blunt arrows indicate promotion and repression of the gene. Yellow
coloured boxes indicate the switched on status of the gene and blue
colour denotes its switched off state.
Petri nets (PN): A unique type of bipartite graph consisting of two
types of nodes: places and transitions. When a transition is activated,
a token is transferred from the input places to the output places; the
process is asynchronous and non-deterministic.
Polynomial dynamical systems (PDS)- An algebraically based approach
that represents a specific type of sequential FDS (Finite Dynamical
System) operating over a finite field. Each transition function is an
element within a polynomial ring defined over the finite field. It
employs advanced rapid techniques from computer algebra and
computational algebraic geometry, originating from the Buchberger
algorithm, to compute the Gröbner bases of ideals in these rings. An
ideal consists of a set of polynomials that remain closed under
polynomial combinations.
Differential equation models (ODE and PDE)- Ordinary Differential
Equations (ODEs) are commonly utilized to represent the temporal
dynamics of networks, while Partial Differential Equations (PDEs) are
employed to describe behaviors occurring in both space and time,
enabling the modeling of pattern formation. These spatiotemporal
Diffusion-Reaction Systems demonstrate the emergence of self-organizing
patterns, typically articulated by the general local activity principle,
which elucidates the factors contributing to complexity and
self-organization observed in nature.
Bayesian models: This kind of model is commonly referred to as
dynamic models. It utilizes a probabilistic approach that enables the
integration of prior knowledge through Bayes' Theorem. A challenge can
arise when determining the direction of an interaction.
Finite State Linear Model (FSML): This model integrates continuous
variables (such as protein concentration) with discrete elements (like
promoter regions that have a limited number of states) in modeling.
Agent-based models (ABM): Initially created within the fields of
social sciences and economics, it models the behavior of individual
agents (such as genes, mRNAs (siRNA, miRNA, lncRNA), proteins, and
transcription factors) and examines how their interactions influence the
larger system, which in this case is the cell.
Rule – based models: In this approach, molecular interactions are
simulated using local rules that can be utilized even in the absence of a
specific network structure, meaning that the step to infer the network
is not required, allowing these network-free methods to avoid the
complex challenges associated with network inference.
Piecewise-linear differential equation models (PLDE): The model is
composed of a piecewise-linear representation of differential equations
using step functions, along with a collection of inequality restrictions
for the parameter values.
A
simple three protein negative feedback loop modeled with mass action
kinetic differential equations. Each protein interaction is described by
a Michaelis–Menten reaction.
Stochastic models: Models utilizing the Gillespie algorithm for
addressing the chemical master equation provide the likelihood that a
particular molecular species will possess a defined molecular population
or concentration at a specified future point in time. The Gillespie method is the most computationally intensive option
available. In cases where the number of molecules is low or when
modeling the effects of molecular crowding is desired, the stochastic
approach is preferred.The
graph demonstrates the enzymatic conversion of cellulose to glucose
over time where red line denoted cellulose and green line denotes
glucose, with key enzymes facilitating the process and their
concentrations changing as the reaction progresses (Time course run in
COPASI). This is a typical kinetic profile for a multi-enzyme hydrolysis
system.
State Space Model (SSM): Linear or non-linear modeling techniques
that utilize an abstract state space along with various algorithms,
which include Bayesian and other statistical methods, autoregressive
models, and Kalman filtering.
Creating biological models
Researchers begin by choosing a biological pathway and diagramming
all of the protein, gene, and/or metabolic pathways. After determining
all of the interactions, mass action kinetics or enzyme kinetic rate laws are used to describe the speed of the reactions in the system. Using mass-conservation, the differential equations
for the biological system can be constructed. Experiments or parameter
fitting can be done to determine the parameter values to use in the differential equations. These parameter values will be the various kinetic constants required
to fully describe the model. This model determines the behavior of
species in biological systems and bring new insight to the specific
activities of systems. Sometimes it is not possible to gather all
reaction rates of a system. Unknown reaction rates are determined by
simulating the model of known parameters and target behavior which
provides possible parameter values.
The use of constraint-based reconstruction and analysis (COBRA)
methods has become popular among systems biologists to simulate and
predict the metabolic phenotypes, using genome-scale models. One of the
methods is the flux balance analysis
(FBA) approach, by which one can study the biochemical networks and
analyze the flow of metabolites through a particular metabolic network,
by optimizing the objective function of interest (e.g. maximizing
biomass production to predict growth).
Applications in system biology
Systems biology, an interdisciplinary field that combines biology,
data analysis, and mathematical modeling, has revolutionized various
sectors, including medicine, agriculture, and environmental science. By
integrating omics data (genomics, proteomics, metabolomics, etc.),
systems biology provides a holistic understanding of complex biological
systems, enabling advancements in drug discovery, crop improvement, and
environmental impact assessment. This response explores the applications
of systems biology across these domains, highlighting both industrial
and academic research contributions. System biology is used in
agriculture to identify the genetic and metabolic components of complex
characteristics through trait dissection. It aids in the comprehension of plant-pathogen interactions in disease resistance. It is utilized in nutritional quality to enhance nutritional content through metabolic engineering.
Cancer
Approaches to cancer systems biology have made it possible to
effectively combine experimental data with computer algorithms and, as
an exception, to apply actionable targeted medicines for the treatment
of cancer. In order to apply innovative cancer systems biology
techniques and boost their effectiveness for customizing new,
individualized cancer treatment modalities, comprehensive multi-omics
data acquired through the sequencing of tumor samples and experimental
model systems will be crucial.
Cancer systems biology has the potential to provide insights into
intratumor heterogeneity and identify therapeutic options. In
particular, enhanced cancer systems biology methods that incorporate not
only multi-omics data from tumors, but also extensive experimental
models derived from patients can assist clinicians in their
decision-making processes, ultimately aiming to address treatment
failures in cancer.
Drug development
Before the 1990s, phenotypic drug discovery formed the foundation of
most research in drug discovery, utilizing cellular and animal disease
models to find drugs without focusing on a specific molecular target.
However, following the completion of the human genome project,
target-based drug discovery has become the predominant approach in
contemporary pharmaceutical research for various reasons. Gene knockout
and transgenic models enable researchers to investigate and gain
insights into the function of targets and the mechanisms by which drugs
operate on a molecular level. Target-based assays lend themselves better
to high-throughput screening, which simplifies the process of
identifying second-generation drugs—those that improve upon
first-in-class drugs in aspects such as potency, selectivity, and
half-life, especially when combined with structure-based drug design. To
do this, researchers utilize the three-dimensional structure of target
proteins and computational models of interactions between small
molecules and those targets to aid in the identification of superior
compounds.
Food safety and quality
The multi-omics technologies in system biology can be also be used in
aspects of food quality and safety. High-throughput omics techniques,
including genomics, proteomics, and metabolomics, offer valuable
insights into the molecular composition of food products, facilitating
the identification of critical elements that affect food quality and
safety. For example, integrating omics data can enhance the
understanding of the metabolic pathways and associated functional gene
patterns that contribute to both the nutritional value and safety of
food crops. This comprehensive approach guarantees the creation of food
products that are both nutritious and safe, capable of satisfying the
increasing global demand.
Environmental system biology
Genomics examines all genes as an evolving system over time,
aiming to understand their interactions and effects on biological
pathways, networks, and physiology in a broader context compared to
genetics. As a result, genomics holds significant potential for discovering
clusters of genes associated with complex disorders, aiding in the
comprehension and management of diseases induced by environmental
factors.
When exploring the interactions between the environment and the
genome as contributors to complex diseases, it is clear that the genome
itself cannot be altered for the time being. However, once these
interactions are recognized, it is feasible to minimize exposure or
adjust lifestyle factors related to the environmental aspect of the
disease. Gene-environment interactions can occur through direct associations
with active metabolites at certain locations within the genome,
potentially leading to mutations that could cause human diseases.
Indirect interactions with the human genome can take place through
intracellular receptors that function as ligand-activated transcription
factors, which modulate gene expression and maintain cellular balance,
or with an environmental factor that may produce detrimental effects. This type of environmental-gene interaction could be more
straightforward to investigate than direct interactions since there are
numerous markers of this kind of interaction that are readily measurable
before the disease manifests. Examples of this include the expression
of cytochrome P450 genes following exposure to environmental substances,
such as the polycyclic aromatic hydrocarbon benzo[a]pyrene, which binds
to the Ah receptor.
Technical challenges
One of the main challenges in systems biology is the connection
between experimental descriptions, observations, data, models, and the
assumptions that stem from them. In essence, systems biology must be
understood within an information management framework that significantly
encompasses experimental life sciences. Models are created using
various languages or representation schemes, each suitable for conveying
and reasoning about distinct sets of characteristics. There is no
single universal language for systems biology that can adequately cover
the diverse phenomena we aim to investigate. However, this intricate
scenario overlooks two important aspects. Models can be developed in
multiple versions over time and by different research teams. Conflicts
can occur, and observations may be disputed. Various researchers might
produce models in different versions and configurations. The
unpredictable elements suggest that systems biology is not likely to
yield a definitive collection of established models. Instead, we can
expect a rich ecosystem of models to develop within a structure that
fosters discussion and cooperation among participants. Challenges also
exist in verifying the constraints and creating modeling frameworks with
robust compositional strategies. This may create a need to handle
models that may conflict with one another, whether between schemes or
across different scales. In the end, the goal could involve the
creation of personalized models that reflect differences in physiology,
as opposed to universal models of biological processes.
Other challenges include the massive amount of data created by
high-throughput omics technologies which presents considerable
challenges in terms of computation and storage. Each analysis in omics
can result in data files ranging from terabytes to petabytes, which
requires strong computational systems and ample storage solutions to
manage and process these datasets effectively. The computational requirements are made more difficult by the necessity
for advanced algorithms that can integrate and analyze diverse,
high-dimensional data. Approaches like deep learning and network-based
methods have displayed potential in tackling these issues, but they also
demand significant computational power.
Artificial intelligence (AI) in systems biology
Utilizing AI in Systems Biology enables scientists to uncover novel
insights into the intricate relationships present within biological
systems, such as those among genes, proteins, and cells.
A significant focus within Systems Biology is the application of AI for
the analysis of expansive and complex datasets, including multi-omics data produced by high-throughput methods like next-generation sequencing and proteomics.
Approaches powered by AI can be employed to detect patterns and
correlations within these datasets and to anticipate the behavior of
biological systems under varying conditions.
For instance, artificial intelligence can identify genes that are
expressed differently across various cancer types or detect small
molecules linked to particular disease states. A key difficulty in analyzing multi-omics
data is the integration of information from multiple sources. AI can
create integrative models that consider the intricate interactions
between different types of molecular data. These models may be utilized
to uncover new biomarkers
or therapeutic targets for diseases, as well as to enhance our
understanding of fundamental biological processes. By significantly
speeding up our comprehension of complex biological systems, AI has the
potential to lead to new treatments and therapies for a range of
diseases.
Structural systems biology is a multidisciplinary field that
merges systems biology with structural biology to investigate biological
systems at the molecular scale. This domain strives for a thorough
understanding of how biological molecules interact and function within
cells, tissues, and organisms. The integration of AI in structural
systems biology has become increasingly vital for examining extensive
and complex datasets and modeling the behavior of biological systems. AI
facilitates the analysis of protein–protein interaction
networks within structural systems biology. These networks can be
explored using graph theory and various mathematical methods, uncovering
key characteristics such as hubs and modules. AI can also assist in the discovery of new drugs or therapies by
predicting the effect of a drug on a particular biological component or
pathway.
The study of romantic love is still in its infancy. As of 2021, there were a total of 42 biological studies on romantic love.
Definition of romantic love
The meaning of the term "romantic love" has changed considerably throughout history, making it difficult to simply define. Initially it was coined to refer to certain attitudes and behaviors described in a body of literature now referred to as courtly love. However, academic psychology and especially biology also consider romantic love in a different sense, which refers to a brain system (or systems) related to pair bonding or mating with associated psychological properties.
Bode and Kushnick undertook a comprehensive review of romantic
love from a biological perspective in 2021. They considered the
psychology of romantic love, its mechanisms, development across the
lifespan, functions, and evolutionary history. Based on the content of that review, they proposed a biological definition of romantic love:
Romantic love is a motivational state
typically associated with a desire for long-term mating with a
particular individual. It occurs across the lifespan and is associated
with distinctive cognitive, emotional, behavioral, social, genetic,
neural, and endocrine activity in both sexes. Throughout much of the
life course, it serves mate choice, courtship, sex,
and pair-bonding functions. It is a suite of adaptations and
by-products that arose sometime during the recent evolutionary history
of humans.
Romantic love in this sense is also not necessarily "dyadic",
"social" or "interpersonal", despite being related to pair bonding.
Romantic love can be experienced outside the context of a relationship, for example in the case of unrequited love where the feelings are not reciprocated. A person can develop romantic love feelings before any relationship has occurred, for only a potential partner. The potential partner can even be somebody they do not know well or are not acquainted with at all, as in cases of love at first sight and parasocial attachments.
The early stage of romantic love (which has obsessive and addictive features) might also be referred to as being "in love", passionate love, infatuation, limerence or obsessive love. Research has never settled on a unified terminology or set of methods. Distinctions are drawn between this early stage of romantic love and the "attachment system" theorized by the attachment theorists like John Bowlby. In the past, attachment theorists have argued that attachment theory
and attachment styles can replace other theories of love, but academics
on love have argued this is incorrect and that romantic love and
attachment are not identical concepts. The early stage of romantic love is thought to involve additional brain
systems for other purposes, with distinct evolutionary histories. Romantic love is also distinct from sexual attraction, although they most often occur together.
Variation exists in the way romantic love is expressed in the
population. A cross-cultural study of currently in-love people found
four clusters, with varying degrees of intensity, obsessive thinking,
commitment, frequency of sex and other differences. Other studies indicate romantic love can be experienced both with or without obsessional features. Typically, intense romantic love is limited to a duration of 12–18 months or as long as 3 years, depending on the estimate; however, in a rare phenomenon called "long-term intense romantic love",
some people experience intense attraction inside a relationship, even
for 10 years or more. This is similar to early-stage intense romantic
love, but at this later stage they exhibit less of the obsessional
features.
Helen Fisher and her colleagues proposed that the brain systems involved with mammalian reproduction can be separated into at least three parts:
Neuroscientists currently believe
that the basic emotions arise from distinct circuits (or systems) of
neural activity; that humans share several of these primary
emotion-motivation circuits with other mammals; and that these brain
systems evolved to direct behavior [...]. It is hypothesized that among
these primary neural systems are at least three discrete, interrelated
emotion-motivation systems in the mammalian brain for mating,
reproduction, and parenting: lust, attraction, and attachment [...].
Attraction (or early-stage romantic love, also called passionate love or infatuation) is associated with feelings of exhilaration, obsessive (or "intrusive") thoughts and the craving for emotional union.
Attachment (the attachment system from attachment theory, and also called companionate love) is associated with feelings of calm, security and comfort, but separation anxiety when apart.
In Fisher's theory, the systems tend to act in unison, but may become
disassociated and act independently. For example, a person in a
long-term partnership may feel deep attachment for their spouse, while experiencing intense romantic love (attraction) for some other individual, while being sexually attracted (lust) to still others, all at the same time. Lisa Diamond
has also used independent emotions theory to explain why people can
'fall in love' sometimes without sexual desire, as in the case of "platonic" infatuation for a friend.
Fisher associates each system with different neurotransmitters and/or hormones (lust: estrogen & androgens; attraction: dopamine, norepinephrine & serotonin; attachment: oxytocin & vasopressin), but modern research shows these associations are not as clearly defined as Fisher's theory proposes. Additionally, romantic love has been associated with endogenous opioids, cortisol and nerve growth factor, which are not included in Fisher's earlier model. With respect to the idea that the systems are independent, a more
modern theory holds that the attachment system is active in early-stage
romantic love, in addition to the infatuation component. Fisher's model
is considered outdated, although the idea of interrelated systems is
useful.
Evolutionary psychology
is seen as an organizing framework which offers explanations behind
psychological functions (rather than merely describing them), as well as
specifying theoretical constraints, like requiring that a given trait
is adaptive in the form of providing reproductive benefit to an individual. Evolutionary psychology has proposed a variety of explanations for romantic love.
Romantic love is a powerful commitment device. Romantic love suppresses the search for alternative mates (even
irrationally so, when a more desirable one comes along), and signals
this to the partner. Romantic love may also signal to alternative mates, disincentivizing them from pursuing oneself. The emergence of longer pair bonds in the evolution of humans coincided with the emergence of concealed ovulation, where it cannot (in general) be determined when a woman is ovulating, requiring partners to stay together while having sex during the entire menstrual cycle. Commitment is seen as adaptive to facilitate this, and to facilitate child care. Love feelings might be the psychological reward produced by the brain when the problem of commitment is being solved.
The intensity of romantic love feelings and why people become "fools for love" can be explained in terms of the handicap principle, which states that a contention arises between honest and fake signaling. When real emotions
evolved, a niche would have been created for sham emotions (e.g. fake
facial expressions) which are less risky to express. One explanation for
why honest signals can evolve without becoming worthless (because of
competing fakers) is that the honest signal can evolve if it is too
expensive to fake. One example in nature is the peacock's tail, an example of conspicuous consumption, a cumbersome display which consumes nutrients. Only a healthy peacock can afford it, so in that case it may have evolved because
it was a handicap, and used by females of the species as an indicator
of health. Romantic love may have evolved to be as bewitching and
besotting as it is, "like handcuffing oneself to railroad tracks", as a
handicap meant to prove that one's commitment is truly real.
Romantic love may have evolved to override rationality, so that
one reproduces regardless of the considerable costs of raising a child,
and regardless of any rational will to be single or child-free.
Romantic love signals parental investment. Paternal investment
in the form of pair bonds has been linked to better outcomes for
children, both as infants and as they grow older. Children raised in
pair bonds are more socially competitive and more likely to survive to
reproductive age.
Being in love makes people more creative, so romantic love may have evolved as a courtship display. It has been suggested that art, music and literature
serve a function like a peacock's tail, but as a display of mental
prowess, designed to impress and make a potential partner swoon.
Creativity is believed by some authors to be especially a part of the
male courtship display.
Romantic love may conserve time and metabolic energy by focusing courtship efforts on a specific individual over others.
Successful pair bonding predicts better health and survival. Happy, well-functioning romantic relationships contribute to mental and
physical health, especially when stress is encountered. The end of a
pair bond (e.g. divorce) is associated with vulnerability, such as to disease, depression, substance abuse, or negative outcomes for children. Victims of a heart attack, for example, are more likely to have another when they live alone.
Monogamous pair bonding helps prevent sexually transmitted diseases (STDs) which compromise fertility, especially for women. Certain STDs (e.g. syphilis) increase the risk of miscarriage,
and otherwise harm or can be passed to an unborn child. The strongest
predictor of contracting an STD is the number of sexual partners, so
limiting this is the best way to limit the risk of contracting a disease
which would harm one's reproductive health.
Romantic love promotes exclusivity via mate guarding. Romantic jealousy
is one of the most common correlates of being in love, which evolved as
a protection from the threat of losing one's love to a romantic rival. Jealousy is seen as adaptive (when it motivates one to maintain their
relationship) up to a point, but can also take the form of pathological jealousy where a sufferer has a delusional or paranoid belief in their partner's infidelity regardless of actual evidence.
Time of evolution
Unsolved problem in biology
When did human pair bonding evolve? Is pair
bonding an antecedent to romantic love, or have there been other steps
in the evolution of pair bonds in humans (e.g. a seasonal bond)?
Although the exact moment during human evolution is unknown, modern romantic love is usually believed to have evolved either during or after the time of bipedalism. The earliest hominid found with extensive evidence of bipedalism (and some evidence of pair bonding) is Ardipithecus ramidus, from about 4.4 million years ago, although it may also be the case that bipedalism is older than this.It has been proposed that monogamous pair bonding (which is rare among mammals)
evolved during this time, because walking bipedally requires mothers to
carry infants in their arms or on their hip, instead of on their backs.
With their hands occupied, mothers would be more vulnerable, requiring
additional help for food and protection from males of the species
(hence, husbands or fathers). A different selection pressure which has been proposed is the evolution of infant altriciality (immaturity and helplessness) and large brain size at birth, which occurred around 2 million years ago. At this time, brain size became so large that a fully-developed infant's head could not fit through the mother's pelvic birth canal (known as the obstetricial dilemma),
requiring the infant to be born early and underdeveloped in comparison
to other species. This would have also placed a greater burden on
mothers, and made paternal support more valuable.
Due to the general scarcity of evidence, it is still also
possible that romantic love (or a precursor to it) predated bipedalism
and altriciality, possibly originating in a common ancestor of humans
and chimpanzees,
5-8 million years ago. While chimpanzees primarily mate
opportunistically, some of their rarer reproductive strategies have
features reminiscent of romantic love (involving mate guarding, and a more-than-fleeting association.) One assumption behind hypotheses based on fossil evidence is that less sexual dimorphism in body mass
(i.e. similarity) is indicative of monogamy, but the comparative
similarity between the sexes in human body mass occurred as recently as
500,000 years ago. This suggests that there may have been multiple steps
in the evolution of human pair bonding, and romantic love may have
evolved during any of these periods.
Courtship attraction
"It was evidently a case of love at first sight, for she swam about the new-comer caressingly...with overtures of affection" (Darwin, 1871, observing a Mallard).
Helen Fisher's
theory is that romantic love (which she considers distinct from
attachment) is a motivation system for choosing and focusing energy on a
preferred mating partner. According to Fisher, this brain system
evolved for mammalian mate choice,
also called "courtship attraction". In this phenomenon, a preferred
mating partner is chosen based on a display of physical traits (such as a
peacock's tail feathers) or other behaviors. Fisher also includes the attraction to personality traits and other characteristics in her mate choice theory for humans. Courtship attraction shares similar behaviors with romantic love in humans, and both involve activation of dopaminergic
reward circuits. In most species, courtship attraction is as brief as
lasting only minutes, hours, days or weeks, but intense romantic love
can last much longer in humans. Fisher believes that during the timeline of human evolution,
mammalian courtship attraction may have become prolonged and
intensified as pair bonding evolved, eventually becoming the phenomenon
of romantic love today.
A critique of Fisher's theory published by Adam Bode holds that courtship attraction only encompasses love at first sight attraction or a crush,
and the core components of romantic love (including the intense
attraction and obsessive thoughts, in addition to attachment) evolved as
a co-option of mother-infant bonding. A study on love at first sight found that even though people reporting the experience retrospectively will recall features resembling passionate love ("constant thoughts about the person and the desire to be with him or her"), people reporting love at first sight currently
after just meeting the potential partner only report neutral scores
(neither agreeing nor disagreeing) on a romantic love measure including a
passion component. Some authors have speculated that the remembered
account of falling in love at first sight (with high passion) is often
actually a memory confabulation. Furthermore, the study found that the experience of love at first sight was related to the physical attractiveness
of the potential partner. This led the researchers to conclude that
love at first sight is actually a strong initial attraction, rather than
resembling the state of being in love. Bode argues this more closely resembles the concept of courtship
attraction, and can be considered a separate system from core romantic
love components. Courtship attraction may be characterized by dopamine, oxytocin and opioid activity, but little is known about it because existing studies were not designed to target it.
Co-option of mother-infant bonding
Co-option is an evolutionary process whereby a given trait is repurposed to take on a new function. One example is how a number of species of fish (e.g. catfish) have co-opted their gas bladder to produce sound. Co-opted traits can be morphological, but also behavioral. Co-option has been used as an explanation of how a species can develop an evolutionary adaptation very quickly sometimes, seemingly faster than Darwinism
could explain. With this process, a seemingly "new" trait can develop
quickly because its structure predated the time of adaptation, only
needing to be modified to function in a new way. In some cases,
co-option involves one gene
whose function is altered, while in other cases the co-opted gene is a
duplicate and the function of the original gene is retained. The terms "co-option" and "exaptation" are closely related, but have different connotations. Exaptation refers to structural continuity when a trait takes on a new function.
Adam Bode has proposed that romantic love is "a suite of
adaptations and by-products" consisting of a number of interrelated
systems, several of which evolved by co-opting mother-infant bonding
(attraction for bonding, obsessive thinking and attachment). The
co-option theory says that the genes that regulate mother-infant bonding
were recreated and took on a new function. Courtship attraction and
sexual desire are "causally linked adjuncts" which were not co-opted,
but were combined and modified in romantic love. The theory is based on
the available human evidence, but also a literature arising from
research on prairie voles that pair bonding uses the same mechanisms that mother-infant bonding uses.[
Academic literature has drawn a parallel between romantic love and the mother-infant dyad since the 1980s, with attachment theorists like Cindy Hazan and Phillip Shaver believing the two share a common biological process. In 1981, Glenn Wilson suggested a close analogy between adult lovers and the kind of infant attachment studied by John Bowlby. In 1999, James Leckman & Linda Mayes compared features of romantic love and early parental love,
finding substantial similarities. Both are altered mental states
featuring preoccupations, exclusivity of focus, a longing for
reciprocity and idealization of the other. The trajectories of both also
share similarities, with preoccupation increasing during courtship (for romantic love) and around the time of birth (for parental love), then diminishing after a relationship is established (for romantic love) or shortly after the postpartum period (for parental love). (The use of "baby talk" by romantic lovers is another "uncanny" similarity.) In 2004, Andreas Bartels and Semir Zeki were the first to compare romantic love and maternal love with fMRI. This comparison looked at areas known to contain high densities of receptors for the attachment hormones oxytocin and vasopressin. Bartels & Zeki found precise overlap in some specific areas including the striatum (putamen, globus pallidus and caudate nucleus) and some overlap in the ventral tegmental area, areas with dopamine
and oxytocin receptors. Each type of love was also associated with
other unique activations. Notably, maternal love involved the periaqueductal gray matter, an area associated with endogenous pain suppression during intense emotional experiences such as childbirth. Two meta-analyses of fMRI experiments have also found similarities between maternal love and romantic love. A 2022 meta-analysis by Shih et al. found that both types of love were
associated with the left ventral tegmental area (more associated with
the pleasurable aspect of reward, or "liking"), while in addition
romantic love also involved the right ventral tegmental area (more
associated with reward "wanting").
In 2003, Lisa Diamond suggested that adult pair bonding is an exaptation of the affectional bond between infants and caregivers, using this to explain phenomena such as romantic friendships and "platonic" infatuations, or i.e. "romantic" passion without sexual desire.Some instances of this are reported by Dorothy Tennov in her study of "limerence" (i.e. love madness, commonly for an unreachable person), in which a younger woman who otherwise considered herself heterosexual would have this type of reaction towards an older woman. Among other examples are schoolgirls falling "violently in love with
each other, and suffering all the pangs of unrequited attachment,
desperate jealousy etc." (historically called a "smash"), and Native American men who seemed to fall in love with each other and form intense, but non-sexual bonds. Helen Fisher's
theory that sexual desire is a separate system from romantic love and
attachment is also given as theoretical evidence. Diamond argues that
romantic love without sexual desire can even happen in contradiction to
one's sexual orientation:
because it would not have been adaptive for a parent to only be able to
bond with an opposite sex child, so the systems must have evolved
independently from sexual orientation. People most often fall in love
because of sexual desire, but Diamond suggests time spent together and
physical touch can serve as a substitute. Diamond believes the
connection between romantic love and sexual desire is "bidirectional" in
that either one can cause the other to occur because of shared oxytocin pathways in the brain.
New model
Based on contentions over evolutionary theories and Fisher's outdated
neurochemical model, Bode has suggested Fisher's model, while useful
and the predominant one for a time, is oversimplified and proposes five
systems:
Courtship attraction is for choosing and focusing energy on a
preferred mating partner and promotes courtship behaviors. It can take
the form of e.g. love at first sight attraction or a crush
and also be intertwined with other forms of attraction, but might not
precede a relationship in all cases. Courtship attraction may be
associated with dopamine, oxytocin and opioids.
Bonding attraction is the type of attraction for pair bond
formation, characterized by a strong desire for proximity, separation
anxiety when apart, exclusivity of focus and heightened awareness of the
loved one. Bonding attraction is associated with dopamine and oxytocin
activity, especially in the ventral tegmental area. According to Bode's arguments, this is the type of romantic attraction shown in fMRI experiments of early-stage romantic love.
Obsessive thinking involves preoccupation or intrusive
thinking about the loved one. Some authors have drawn a comparison
between this feature and obsessive-compulsive disorder, suggesting they share similar neurobiology, but the evidence for that is limited and ambiguous.
Attachment is for pair bond maintenance, or maintaining very
close personal relationships, with psychological features like a
heightened sense of responsibility, longing for reciprocity and a
powerful sense of empathy. Attachment is associated with oxytocin,
dopamine and opioid activity, but there is also some evidence for the
involvement of vasopressin.
Bode suggests that the systems of bonding attraction, obsessive
thinking and attachment (the three systems which were co-opted from
mother-infant bonding) together form the core of romantic love (the
necessary components). However, all five systems are merged into one
single phenomenon of romantic love, with a variety of different outcomes
depending on the circumstances.
The early stage of romantic love is being compared to a behavioral addiction (i.e. addiction to a non-substance) but the "substance" involved is the loved person. Addiction involves a phenomenon known as incentive salience, also called "wanting" (in quotes). This is the property by which cues in the environment stand out to a
person and become attention-grabbing and attractive, like a
"motivational magnet" which pulls a person towards a particular reward. Incentive salience differs from craving in that craving is a conscious
experience and incentive salience may or may not be. While incentive
salience can give feelings of strong urgency to cravings, it can also
motivate behavior unconsciously, as in an experiment where cocaine users were unaware of their own decisions to choose a low dose of cocaine (which they believed was placebo) more often than an actual placebo. In the incentive-sensitization theory
of addiction, repeated drug use renders the brain hypersensitive to
drugs and drug cues, resulting in pathological levels of "wanting" to
use drugs. People in love are thought to experience incentive salience in response
to their beloved. Lovers share other similarities with addicts as well,
like tolerance, dependence, withdrawal, relapse, craving and mood modification.
Incentive salience is mediated by dopamine projections in the mesocorticolimbic pathway of the brain, an area generally involved with reward, motivation and reinforcement learning. Dopamine signaling for incentive salience originates in the ventral tegmental area (VTA) and projects to areas such as the nucleus accumbens (NAc) in the ventral striatum. The VTA is one of two main areas of the brain with neurons which produce dopamine (the other being the substantia nigra pars compacta). Projections from the VTA innervate the NAc, where dopamine activity attaches motivational significance to stimuli associated with rewards. Brain scans of people in love using fMRI (commonly while looking at a photograph of their beloved) show activations in these areas like the VTA and NAc. Another dopamine-rich area of the reward system shown to be active in romantic love is the caudate nucleus, containing 80% of the brain's dopamine receptor sites, located in the dorsal striatum. The dorsal striatum is implicated in reinforcement learning, and the caudate nucleus has shown activity in response to a monetary reward and cocaine. This activity in reward and motivation areas suggests that early-stage
intense romantic love is a motivation system or goal-oriented state
(rather than a specific emotion), consistent with the description of
romantic love as a desire or longing for union with another person. These activations are also consistent with the similarity between romantic love and addiction.
In addiction research, a distinction is drawn between "wanting" a
reward (i.e. incentive salience, tied to mesocorticolimbic dopamine)
and "liking" a reward (i.e. pleasure, tied to hedonic hotspots), aspects which are dissociable. People can be addicted to drugs and compulsively seek them out, even when taking the drug no longer results in a high or the addiction is detrimental to one's life. They can also irrationally "want" (i.e. feel compelled towards, in the sense of incentive salience) something which they do not cognitively wish for. In a similar way, people who are in love may "want" a loved person even
when interactions with them are not pleasurable. For example, they may
want to contact an ex-partner after a rejection, even when the
experience will only be painful. It is also possible for a person to be "in love" with somebody they do not like, or who treats them poorly.
Academics have proposed a number of theories for how addictions begin and perpetuate. One prominent theory developed by Wolfram Schultz involves a dopamine signal which encodes a reward prediction error (RPE): the difference between the predicted value of a reward which would be received by performing a particular action and the actual value upon receiving it (i.e. whether the reward was better, equal to or worse than expected). In this theory, dopamine is also part of a mechanism for reinforcement
learning which associates rewards with the cues which predicted them.
Drugs of abuse like cocaine
hijack this mechanism by artificially overstimulating dopamine neurons,
mimicking an RPE signal which is much stronger than could be produced
naturally. An alternative model developed by Kent Berridge and Terry Robinson
states that dopamine signaling causes the motivational output
(incentive salience) which is proportional to RPE, but that the dopamine
signal itself may be an effect of learning rather than causing it
directly. There is, however, said to be overwhelming evidence that dopamine guides learning in addition to motivation. The computation of dopamine signaling is complicated, with inputs from a
variety of areas in the brain, although its output (primarily from the
VTA) is a relatively homogeneous signal encoding the level of RPE. One study has investigated whether people in long-term romantic
relationships experienced RPE in response to having expectations about
their partners' appraisal of them validated or violated, indicating they
do. This study used fMRI to find that reward areas like the VTA and
striatum responded in a way consistent with other research on RPE. Most fMRI studies of romantic love have had participants look at a
photograph, and the resulting reward system activity has been
interpreted in terms of salience.
Research has not investigated whether romantic love shares all of the neurobiological aspects of addiction. Despite similarities, there are also differences between romantic love and addiction. One of the major differences is that the trajectories diverge, with the
addictive aspects tending to disappear over time during a relationship
in romantic love. By comparison, in a drug addiction, the detrimental
aspects magnify over time with repeated drug use, turning into
compulsions, a loss of control and a negative emotional state. It has been speculated that the difference between these could be related to oxytocin activity present in romantic love, but not in addiction. Oxytocin seems to ameliorate the effects of drug withdrawal, and it
might inhibit the more long-term, excessive effects of addiction.
Academics do not universally agree on whether or not love is always an addiction or when it needs to be treated. The term "love addiction" has had an amorphous definition over the years and does not yet denote a psychiatric condition,
but recently one definition has been developed that "Individuals
addicted to love tend to experience negative moods and affects when away
from their partners and have the strong urge and craving to see their
partner as a way of coping with stressful situations." Other authors include rejected lovers as love addicts, or specify that love is an addiction when it involves abnormal
processes which carry negative consequences. A broader view is that all
love is addiction, or simply an appetite, similar to how humans are dependent on food. Research on behavioral addictions is more limited than research on
drugs of abuse; however, there is a growing body of evidence that some
people are susceptible to showing brain patterns in response to natural
rewards (food, sex, etc.) similar to drug addicts, particularly in the
case of gambling addiction. Romantic love may be a "natural" addiction, which differs from the
nature of drug addiction in that love may be prosocial and has been
evolved for the purpose of pair bonding. Helen Fisher, Arthur Aron
and colleagues have proposed that romantic love is a "positive
addiction" (i.e. not harmful) when requited and a "negative addiction"
when unrequited or inappropriate.
Oxytocin
is sometimes called the "love hormone", because of its involvement in
the mechanisms of maternal behavior and adult pair bonding. Oxytocin is
synthesized primarily in an area of the hypothalamus and released into the blood via the pituitary gland, where it has been found circulating in people in the early stages of romantic love. Additionally, the hypothalamus projects oxytocin to other areas of the brain, like the ventral tegmental area (VTA), nucleus accumbens (NAc), amygdala and hippocampus. The projections to reward areas (VTA and NAc) are thought to modulate social salience, or i.e. the level of dopamine activity in response to socially-relevant stimuli. This oxytocin signaling in reward pathways may also be the source of salience in response to a loved one.
A placebo-controlled study
found that administering intranasal oxytocin enhanced facial
attractiveness of a romantic partner while viewing a photograph, as
compared to an unfamiliar face. The effect was also measured using fMRI, which found enhanced brain activity in reward areas like the VTA and NAc. Another fMRI study found dopamine-rich genetic expression of an oxytocin receptor gene in the left VTA, and the left VTA has also been found active in response to facial attractiveness. In humans, circulating oxytocin levels have been associated with higher
levels of interaction between partners, and also predicted which
couples would still be together 6 months later. Anna Machin calls the combination of oxytocin and dopamine the "glue" which makes the early stages of a relationship possible.
The role of oxytocin in human behavior is varied and complex. Oxytocin lowers inhibitions to forming new relationships by deactivating the amygdala, involved with processing fear and anxiety. Oxytocin can be released with physical touch, hence it's also sometimes called the "cuddle hormone".Oxytocin also plays a role in sexual behavior, being released during sexual arousal and orgasm. Aside from romantic and parental bonding, oxytocin activity has a role
in the interactions with peers or strangers, for example facilitating facial recognition and eye contact. Oxytocin is believed to facilitate trust and altruistic behaviors towards in-groups (e.g. partners or children), but also aggression towards out-groups (e.g. strangers or conspecifics).
Much of the research on oxytocin comes from experiments on monogamousprairie voles (notably by Larry Young), but this research is also used for making inferences about humans. In prairie voles, both oxytocin and dopamine signaling have been shown
to influence pair bond development. For example, the number of oxytocin
receptors in the NAc is positively related to how fast a partner
preference is formed. A partner preference can also be prevented by
injecting either a dopamine or oxytocin receptor antagonist (a drug which blocks transmission) into the NAc of a prairie vole directly.
In a contemporary model of the brain systems involved with
romantic love, this type of salience (or 'bonding attraction') is
present throughout the entire time a person is experiencing romantic
love, including during the early stages. This contrasts with some
previous theories (e.g. proposed by Helen Fisher
in 1998) which stated that oxytocin activity and dopamine activity were
distinct (and independent) systems, and that oxytocin activity only
became prominent at some later stage of a relationship. Levels of oxytocin would still vary from situation to situation because
of differing types of stimuli, for example because of less regular
interaction and physical touch in cases of unrequited love. This could be used to explain some of the maladaptive symptoms of infatuation (e.g. sleep difficulties, social anxiety, clammy hands, etc.), when dopaminergic activity is high without the calming effect of oxytocin from the attachment system.
Modern research is increasingly showing the importance of endogenous opioids in love and social attachment, particularly the β-endorphin (the most potent endogenous opioid) and the μ-opioid receptor system. While opioids have their origin being the body's natural painkiller, they're also implicated in a variety of other systems, essentially like neurotransmitters. Opioid receptors are located throughout the brain, including in the limbic system (affecting basic emotions) and neocortex (affecting more conscious decision-making). Opioids are linked to the consummatory part of reward, or i.e. "liking" or pleasure, and released in areas of the brain called hedonic hotspots (or pleasure centers). Hedonic hotspots are located in the nucleus accumbens, the ventral pallidum and other areas.This function includes social reward, or the pleasurable aspect of social interactions.
The brain opioid theory of social attachment (BOTSA) is a
long-running theory summarizing this connection, originally formulated
in the 1980s and 1990s, based on a proposal by the psychiatristMichael Liebowitz and research by the neuroscientistJaak Panksepp. Starting in the 1990s, opioids were overshadowed by the interest in oxytocin and largely overlooked until more recently, possibly because of the difficulty studying them (requiring e.g. a PET scan, which is expensive). Opioids have been connected to a variety of social experiences, including the early stage of romantic love and attachment styles. While the addictive aspects of love have been compared to cocaine or amphetamine addiction, other aspects may also resemble an opioid addiction.
BOTSA (as it was originally conceived of) predicts that in the absence of social relationships,
individuals will have comparatively lower levels of endogenous opioids,
motivating them to initiate contact with other people. Social contact
then leads to feelings of euphoria and contentment, but individuals also need to continue contact to avoid withdrawal symptoms. Liebowitz originally argued that romantic relationships resemble narcotic addiction, and that individual neurochemical differences could also explain why some people are unable to commit, or stay in abusive relationships. Earlier experiments on BOTSA were animal studies, but in the 2000s this has been expanded to include human experiments.
Among the animal studies which have been done, there is some
evidence that separation distress is akin to opioid withdrawal. Studies
on chicks, puppies, Guinea pigs, rats, sheep and monkeys have shown that administrating morphine reduces distress vocalizations when separated from the mother, and administrating naloxone (an opioid antagonist) increases them, even in the presence of other members of the same species. In another study, mutant mouse
pups with a μ-opioid receptor knockout (lacking the μ-receptor gene)
vocalized less frequently in response to isolation than normal mice.
Administration of morphine had no effect on distress vocalization
frequency in the knockout mice, despite reducing it in normal mice.
Furthermore, these knockout mice had a reduced preference for their
mother's odor, which is normally the result of conditioning mediated by
the endogenous opioid system. In nonhuman primates, studies have suggested that endogenous opioids provide the euphoria behind dyadic social grooming behaviors. Other animal studies have shown that endogenous opioids play a role in the desire for rough-and-tumble play (a physical, but also social behavior). In humans, physical activity with a social element (rowing, dancing, laughing) increased pain tolerance more when the activities were synchronized with other people.
An fMRI
experiment in 2010 investigated whether viewing a picture of a romantic
partner could reduce pain sensitivity, and which areas of the brain
became active. Participants were exposed to high temperatures (resulting
in moderate or high pain levels) while viewing a picture of a romantic
partner (whom they were intensely in love with), or a friend, or
performing a word association task which has also been shown to reduce pain via distraction. Participants were then asked to rate how much pain they felt on a pain scale,
and both viewing a romantic partner and performing the distraction task
(but not viewing a friend) were found to reduce pain levels. The fMRI
scans revealed that viewing a romantic partner activated reward circuits
in the brain, while the distraction task did not. Brain areas were also
correlated with pain relief to reveal that reward analgesia and
distraction analgesia involved distinct areas. Some areas associated
with sensory processing of pain also had decreased activity while viewing a romantic partner. An earlier experiment showed that viewing photos of a romantic partner
reduced experimental pain, but did not pair it with a brain scan.
A PET scan experiment in 2016 investigated whether non-sexual
social touching between romantic partners was mediated by endogenous
opioid activity. This study found that social touch did have an effect,
but unexpectedly found that social touching decreased opioid activity in
the brain rather than increasing it (despite being rated as pleasurable
by participants). This is in contrast with prior PET research that
pleasant affect is related to increased opioid activity. One possible
explanation is that touching decreases stress, so this might also
decrease the ongoing opioid activity in response to distress and pain.
As this is also at odds (to some extent) with primate studies on
grooming, there may be some variation between species in how opioids are
involved with social reward. Other modern studies on humans include blood plasma levels, genetics and studies with drugs like morphine and naltrexone to see how they change social perception and behavior.
Obsessive thinking
Unlike OCD, passionate love (as in limerence) starts with a period of intoxicating joy, and only later reaches a state of anxiety when unrequited. The thoughts also differ in function and content.In addiction, the early stage starts with positive reinforcement (binging and intoxication), but over time a transition occurs towards a more compulsive stage of negative reinforcement (avoiding withdrawal). This later stage (of negative reinforcement) is a possible parallel with OCD (where compulsions relieve tension or anxiety).
Obsessive thinking about a loved one has been called a hallmark or a cardinal trait of romantic love, ensuring that the loved one is not forgotten. Some reports have been made that people can even spend as much as 85 to
100% of their days and nights thinking about a love object. One study found that on average people in love spent 65% of their waking hours thinking about their beloved. Another study used cluster analysis
to find several different groups of lovers, with the least intense
group spending 35% of their time on average and the most intense at 72%. Since the late 1990s, these obsessional features have been compared to obsessive–compulsive disorder (OCD). This is also sometimes paired with a theory that obsessive (or intrusive) thinking is related to serotonin levels being lowered while in love, although study results have been inconsistent or negative. Another theory relates obsessive thinking to addiction, because drug users exhibit obsessive thoughts about drug use, as well as compulsions.
In 1999, James Leckman
and Linda Mayes published a theoretical comparison between early-stage
romantic love, early parental love and OCD. This paper was intended as
an investigation into the origin of OCD, but it also relates to the
evolutionary theory of romantic love. Both early-stage romantic love and OCD share features of preoccupation,
intrusive thoughts, a heightened sense of responsibility, a need for
things to be "just right" and some proximity-seeking behaviors. In some
cases, obsessions experienced by OCD patients relate to what harms might
happen to a family member, which resembles some behavioral patterns
involved with romantic and parental love. The authors also speculate
that psychasthenia
(feelings of incompleteness, insufficiency or imperfection) resembles
the "longing for reciprocity" and idealization which are features of
romantic love.
Two experiments have investigated whether there is a relationship
between romantic love and serotonin levels, by taking different
measures using blood samples. Although serotonin levels in the central nervous system
would actually be the measure of interest, it has been assumed that
measures of peripheral serotonin can be used as a marker for this. A 1999 experiment led by Donatella Marazziti found that people in love had plateletserotonin transporter
(SERT) density which was lower than controls, and similar to the
density of a group of unmedicated OCD patients. Six of the 20 in-love
participants were also retested after a period of 12 to 18 months, and
SERT density had returned to normal. However, because Marazziti's experiment looked at SERT (rather than
serotonin directly), this makes it ambiguous whether serotonin levels
were actually higher or lower. SERT transports serotonin from blood plasma back into the platelets, so that a reduction in SERT could correspond to an increased plasma level.
Another experiment in 2012 led by Sandra Langeslag
which looked at blood serotonin levels found a contradictory result,
with men and women being affected differently. Men had lower serotonin
levels than controls, but women had higher serotonin levels. In women,
obsessive thinking was also actually associated with increased
serotonin. A 2025 study led by Adam Bode also found no association between SSRI
use and obsessive thinking about a loved one, or the intensity of
romantic love. Therefore, although the earlier experiments do suggest
romantic love and serotonin are probably associated, the authors suggest
that the idea of obsessive thinking being attributed to lowered
serotonin levels seems inaccurate.
Emotional valence
Rather than being a specific emotion itself, romantic love is believed to be a motivation
or drive which elicits different emotions depending on the situation:
positive feelings when things go well, and negative feelings when awry. Reciprocated love may elicit feelings of joy, ecstasy, or fulfillment, for example, but unrequited love may elicit feelings of sadness, anxiety, or despair. A 2014 study of Iranian young adults found that the early stage of romantic love was associated with the brighter side of hypomania (elation, mental and physical activity, and positive social interaction) and better sleep quality, but also stronger symptoms of depression and anxiety. Those authors conclude that romantic love is "not entirely a joyful and happy period of life". Romantic love may be either pleasant or unpleasant, regardless of the intensity level. One of Dorothy Tennov's
interview participants recalls being in love this way: "When I felt
[Barry] loved me, I was intensely in love and deliriously happy; when he
seemed rejecting, I was still intensely in love, only miserable beyond
words." The intensity of love feelings is also distinct from whether an
individual is satisfied with their relationship (although the measures
have been shown to be related to some extent). One can be satisfied with
their relationship because it fulfills some other need besides love for
their partner (like money or child care), or conversely be in love with an abuser in an abusive relationship.
Unrequited love is common among young adults. A study by Roy Baumeister
and Sarah Stillwell found that 92.8% of participants reported at least
one "powerful or moderate" experience of unrequited love in the past 5
years. A different study found 63% had a "huge crush" at least once in the
past 2 years (but not letting the person know), and unrequited love was
four times more frequent than equal love. Another found that 20% had experienced unrequited love more than 5
times, according to a definition that "When one is experiencing this
emotion, it has been described as having one’s emotions on a roller
coaster, finding it difficult to concentrate, and thinking constantly
about the person with whom you are in love. The person is said to have
the power to produce extreme highs and lows of emotion in you, depending
on how he or she acts towards you."
In 2010, Helen Fisher, Arthur Aron and colleagues published their fMRI
experiment investigating which areas of the brain might be active in
recently rejected lovers. Participants had been in a relationship with
their ex-partner for an average of 21 months, and then were post-rejection for an average of 63 days at the time of the experiment. These participants reported spending more than 85% of their waking hours thinking of their rejector, reported a lack of emotional control, and exhibited unhappiness, with sometimes more extreme emotions like depression, anger, and even paranoia in pre- and post-interviews.Similar to other fMRI experiments, the scan while looking at a photograph of the rejecting partner showed activations in dopaminergicreward system areas, like the ventral tegmental area and nucleus accumbens.
These activations were also stronger than in a previous experiment of
participants who were happily in love. The nucleus accumbens, prefrontal cortex and orbitofrontal cortex which were active have been associated with assessing one's gains and losses, and areas of the insular cortex and anterior cingulate cortex which were active have been involved with physical pain and pain regulation (respectively) in other studies.
Stress and physiological arousal
In the early stages of romantic love, individuals may start out hypervigilant
(hyperaware and sensitive to a partner's cues) due to uncertainty and
novelty, but become synchronized over time as a relationship progresses.
Bonding is thought to be in part facilitated by coordinated behaviors
which display reciprocity and events which evoke beneficial stress (eustress), like a passionate kiss. The stress response system involves two major systems: the autonomic nervous system and the hypothalamic–pituitary–adrenal axis (HPA axis). Some experiments have been done which support the idea that the stress
response is involved during the early stage of romantic love, measuring cortisol levels; however, these experiments have been inconsistent with respect to cortisol being higher or lower.
In drug addiction, corticotropin release factor (CRF) is involved with the aversive effects of withdrawal. Stress causes CRF to release into the ventral tegmental area and nucleus accumbens
shell, motivating reinstatement of drug use. A similar effect is also
hypothesized in pair bonds, where stress after separation or social loss
motivates a person to return to the partner; however, experiments have
not investigated this in humans, only rodents.
Helen Fisher believed that separation anxiety activates the HPA axis, producing these stress hormones. It's ironic, she says, because short-term stress can also produce dopamine and norepinephrine, so "as the adored one slips away, the very chemicals that contribute to feelings of romance grow even more potent".
Frustration attraction and uncertainty
Dopamineneurons in the ventral tegmental area
are theorized to encode a "reward prediction error" (RPE) signal,
rather than a reward per se. This RPE signaling indicates whether a
given reward was either better, equal to, or worse than what was
anticipated, and this is believed to be part of a reinforcement learning paradigm. Studies have shown that for learning about a stimulus to occur (so that
behavior in response to it changes), the reward has to be surprising or
unpredicted. Rewards which are better than predicted reinforce the behavior and
cause it to become more frequent, while a reward which is worse than
expected would be avoided. Dopamine neurons increase their firing rate when encountering an unexpected reward. After reinforcement learning occurs, dopamine neurons also fire in
response to encountering cues in the environment which predicted the
reward (e.g. in animal studies, a lever or a special sound). As predictions become updated and the rewards are the same as expected, dopamine activity comparatively diminishes.
"Frustration attraction" (also called the "Romeo and Juliet effect") is the idea that adversity heightens romantic passion, for example, through social or physical barriers. The phenomenon has been remarked on by many authors, such as Socrates, Ovid, the Kama Sutra, and "Dear Abby". Bertrand Russell
once opined that "when a man has no difficulty in obtaining a woman,
his feeling toward her does not take the form of romantic love". Some common social barriers are parents who interfere with their children's romance (as in Romeo and Juliet), deceived spouses, or other social customs. Helen Fisher
believes the phenomenon can be explained by the mechanics of dopamine,
because animal studies have shown that when a reward which is
anticipated to be incoming is delayed, reward-expecting neurons prolong
their firing (over comparatively short timescales—in these studies)
until the reward is delivered.
Passionate or infatuated love is also said to thrive in situations which involve the uncertainty of intermittent reinforcement,
when consummation is withheld, when barriers prevent lovers from
meeting regularly, or when one's perceptions of how likely their love is
reciprocated are ambiguous and constantly changing. Uncertainty seems to magnify cue-triggered incentive salience "wanting". A comparable type of situation is that of a slot machine, where the rewards are designed to be always unpredictable so the gambler cannot understand the pattern. Unable to habituate to the experience, for some people the exhilarating high from the unexpected wins leads to gambling addiction and compulsions. If the machine paid out on a regular interval (so that the rewards were expected), it would not be as exciting.
Uncertainty theory in the context of romantic love is associated with Dorothy Tennov's theory of limerence, an addictive, infatuated kind of love, commonly experienced for an unobtainable or unreachable person. In her study, Tennov observed reports of sometimes drastic emotional
transitions caused by changes in one's perception over whether their
love might be reciprocated, and these abrupt transitions could cause
seeming emotional volatility even in otherwise stable individuals. The effect of uncertainty has also been interpreted as attachment anxiety.
Intermittent maltreatment (known as "traumatic bonding" in abusive relationships)
is also believed to intensify romantic "passion" (i.e. strong emotion,
including suffering). This is, again, believed to be related to
intermittent reinforcement and how one's expectations are confirmed or
violated.According to Elaine Hatfield,
'Consistency generates little emotion; it is inconsistency that we
respond to. If a person always treats us with love and respect, we start
to take that person for granted. We like him or her—but "ho hum." [...]
What would generate a spark of interest, however, is if our admiring
friend suddenly started treating us with contempt—or if our arch enemy
started inundating us with kindness.'
Positive illusions
"Crystallization" was coined by the 19th-century French writer Stendhal to refer to these positive illusions, based on an analogy where a tree branch is tossed into a salt mine. The tree branch (or twig) becomes covered in salt crystals, transforming it "into an object of shimmering beauty".
People in love tend to overemphasize the positive aspects of their
loved one or relationship, while overlooking or devaluing negative
aspects. This is regarded as a type of cognitive bias called positive illusions. The phenomenon has also been referred to as crystallization, idealization, "love is blind" bias, putting the loved one on a pedestal, or seeing through rose-colored glasses. In the past, some authors have depicted the phenomenon as a malady,
arguing that people who idealize would have their partner fall short of
their high expectations as a relationship progresses; however, despite
this, significant modern scientific evidence has shown that positive
illusions actually contribute to relationship satisfaction, long-term
well-being and decreased risk for relationship discontinuation.
The exact mechanism is not currently understood, but some brain areas are proposed to be related. The dopaminergic areas of the reward system which are active in romantic love may be involved with attributing salience to the positive characteristics of a loved one. The dorsal anterior cingulate cortex
is involved with error detection and has been active during negative
social evaluation and exclusion, so that reduced activation of this area
would be an adaptive response to a partner's negative characteristics.
Certain areas of the prefrontal cortex
could also be exerting top-down control to suppress emotional responses
to attractive alternatives. Information is then passed to the orbitofrontal cortex, where positive and negative information is weighed, resulting in a biased subjective value about the partner.
Brain imaging
Brain imaging techniques such as functional magnetic resonance imaging (fMRI) and positron emission tomography
(PET) have been used to investigate which brain regions are involved in
romantic love. Nearly all of these experiments have had participants
look at a photograph
of their beloved during an fMRI scan, with a few exceptions, although
the specific procedures used have not always been identical. The
differences in experimental design (e.g. length of time the participants
had been in love, or the specific task given to participants during the
scan) can be used to explain why the experiment results are sometimes
different.
In 2000, a study by Andreas Bartels and Semir Zeki of University College London was the first fMRI study of romantic love. The 17 participants were "truly, deeply and madly in love", had been
together for a mean of 2.4 years, and were shown either one or two
photographs of their loved one during the scan. Two main areas were
active in this study: the middle insular cortex,
associated with stomach churning or "gut feelings", which could have
something to do with the feeling of "butterflies in the stomach", and
part of the anterior cingulate cortex, associated with feelings of euphoria. Other activations were areas in the cerebrum, the caudate nucleus, putamen and the cerebellum. A later analysis in 2004 by the same authors also reports activity in the ventral tegmental area (VTA), which produces dopamine. The study also showed key deactivations, areas of the brain that were
less active in romantic love compared to friendship love, in the amygdala and medial prefrontal cortex (mPFC). The amygdala is involved with fear and risk detection, and the mPFC is involved with understanding and predicting the intentions of other people, called mentalizing.
These deactivations are taken as evidence that "love is blind", or i.e.
that people in love discount the risks involved and misunderstand
people's intentions, even leading to folly sometimes.
In 2005, a study by Arthur Aron, Helen Fisher, Debra Mashek, Greg Strong, Haifang Li and Lucy Brown was the first fMRI study of early-stage intense romantic love.
It has been praised as advancing the scientific understanding of infatuated love, even by a skeptic of fMRI literature. This study differed from Bartels & Zeki in that the 17 participants
who had "just fallen madly in love" had been in love for a much shorter
mean time of only 7.4 months. These participants were more intensely in
love, and spent 85% or more of their waking hours thinking of their
loved one. This study also had participants look at a photograph of their loved
one during the scan. Reward and motivation areas were active, like the
VTA and areas of the caudate. Activity was also found in the insular and
cingulate cortex, involved with emotion. Some interesting areas were correlated with the length of the relationship, like the ventral pallidum, implicated in attachment in prairie voles,
and the anterior cingulate, implicated in obsessive thinking, cognition
and emotion. This study also examined correlations with facial
attractiveness to determine that the right VTA was active because of
romantic passion rather than because the partner was aesthetically
pleasing. Aesthetically pleasing faces elicited more activity in the
left VTA, which is more associated with "liking" a reward (i.e.
pleasure), whereas the right VTA is more associated with "wanting" a
reward (i.e. incentive salience). In 2011, Xu et al. repeated the experiment by Aron et al., but using Chinese participants.
Ortigue et al. used fMRI to investigate the subliminal
influence of romantic love on motivation, interested in how these
implicit neural representations might differ from previous experiments
where subjects were consciously aware of the stimulus (viewing a
photograph). In Ortique et al.'s study, participants were shown a subliminal prime
word for 26ms (either their beloved's name, the name of a friend, or a
word describing a personal passion like a hobby), followed by a series
of symbols (#) for 150ms, followed by a target word for 26ms. This
target was either an English word, non-word or blank, and participants
were asked to identify whether it was a word or not. In trials with the
love prime or passion prime, participants were faster to identify
whether the target was a word or not, and this also correlated with
scores on the Passionate Love Scale.
The authors believe this shows that love priming activates motivation
systems in the brain, rather than just evoking a particular emotion. The fMRI scanning showed brain regions active for love primes similar
to previous experiments, including reward and motivation areas like the
VTA and caudate, but with some additions. Subliminal love priming
additionally activated the bilateral fusiform gyri and angular gyri, involved in integrating abstract representations. The authors relate this to the self-expansion model
of interpersonal relationships, where self-expansion by integrating the
characteristics of one's beloved into one's self (called inclusion of
the other in the self) is a rewarding experience which may promote
romantic love feelings.
In brain scans of long-term intense romantic love (involving
subjects who professed to be "madly" in love, but were together with
their partner 10 years or more) led by Bianca Acevedo,
attraction similar to early-stage romantic love was associated with
dopamine reward center activity ("wanting"), but long-term attachment was associated with the globus palludus, a site for opiate
receptors identified as a hedonic hotspot ("liking"). Long-term
romantic lovers also showed lower levels of obsession compared to those
in the early stage.
An fMRI study led by Sandra Langeslag
investigated the effect of attention on brain activity related to a
loved one. In most other previous experiments, subjects only passively
viewed a photograph, but this experiment used an oddball task
to distinguish between instances where the loved one was either the
intended target of the subject's attention or a distraction.
Participants were given trials where they were presented with a random
face for only 250ms (usually an unknown person) and instructed to watch
for either a loved one or a friend, then press a button if the face was
the intended target for a given run. In some runs, the loved one would
be the intended target for a button press, while the friend would be a
distractor causing participants to press the button by mistake
sometimes, while in other runs the friend would be the target and the
loved one a distractor. This experiment found that activity in the dorsal striatum
(an area of the reward system) was modulated by whether or not
participants were instructed to pay attention to their loved one. That
is, the dorsal striatum showed more response to the loved one than to
the friend, but only when the loved one was the target. This led the
authors to conclude that "the dorsal striatum is not activated by
beloved-related information per se, but only by beloved-related
information that is attended". This activity also tended to be smaller
when participants had been in love or been in a relationship for longer.
The dorsal striatum is implicated in reinforcement
learning, so the authors interpret the increase in brain activity as
reflecting prior reinforcement of social actions which leads the
infatuated individuals to pay preferential attention to their loved one.
Participants also tended to press the button by mistake more often when
distracted by the loved one than the friend.
.jpg)
