Physics and Applied Physics

1. The Field of Study at a Glance

A soap film’s changing colors become a question about how light waves, film thickness, and viewing angle interact, rather than simply an occasion to name the colors.

2. History and the People Who Shaped the Field

Physics developed from astronomy, mathematics, natural philosophy, and instrument making across societies. Ibn al-Haytham’s eleventh-century Optics joined geometrical reasoning with investigations of light and vision. Its later Latin transmission helped connect Arabic and European optical traditions, a history discussed in the University of London Press edition of Abdelhamid I. Sabra’s translation of Books IV–V.

Isaac Newton’s Principia (1687) connected terrestrial and planetary motion through mathematical laws. Oxford’s Newton Project documents exchanges with Edmond Halley and Robert Hooke surrounding its development. Nineteenth-century problems involving electricity, communications, and heat strengthened connections among theory, laboratories, and industry. James Clerk Maxwell’s A Dynamical Theory of the Electromagnetic Field (1865) treated light within an electromagnetic framework.

Albert Einstein’s On the Electrodynamics of Moving Bodies (1905) reworked the relationship between space, time, and motion. Quantum theory subsequently emerged through collective attempts to explain radiation and matter. In 1924, Satyendra Nath Bose developed a new way of counting light quanta; Einstein extended the approach to matter. Kameshwar C. Wali’s 2005 historical essay for the American Physical Society reproduces their correspondence. Emmy Noether’s 1918 work connected continuous symmetries with conservation laws, turning a recurring physical pattern into a precise mathematical relationship.

Applied physics also became an explicit institutional bridge. Cornell’s departmental history dates its Engineering Physics undergraduate program to 1946 and its graduate program in Applied Physics to 1967. These program structures are much newer than practical uses of physics. Classical models remain indispensable, but relativity, quantum theory, and subsequent experiments established limits to assumptions once treated as universal.

3. Main Branches and Areas of Specialization

Physics is organized partly by governing frameworks and partly by the systems being investigated. Common areas include:

• Mechanics and fluids: motion, forces, oscillations, deformation, and flow, from pendulums to collective fluid patterns.
• Electromagnetism, optics, and photonics: fields, radiation, and light–matter interaction, including imaging, waveguides, and lasers.
• Thermal and statistical physics: temperature, energy transfer, entropy, collective behavior, and phase transitions.
• Quantum, atomic, nuclear, and particle physics: quantum states, constituents of matter, and interactions across different energy scales.
• Condensed matter and materials physics: electronic, magnetic, and structural behavior of solids and other many-particle systems.
• Relativity, gravitation, and astrophysics: space-time, high-speed motion, gravity, and physical explanations of cosmic systems.

Applied physics crosses these boundaries through instrumentation, devices, energy, plasmas, nanoscience, and biophysics. Columbia’s 2026–27 Applied Physics B.S. curriculum, for example, lists application areas spanning optics, plasmas, fluids, solid-state physics, and biophysical modeling. Theory, experiment, and computation operate across these areas; they are not mutually exclusive tracks.

4. The Central Questions of the Field

• What belongs inside the system? Boundaries determine which forces, energy flows, and environmental influences must be considered.
• Which quantities remain unchanged, and why? Symmetry offers powerful explanations, but proposed symmetries require scrutiny.
• How does collective behavior emerge? Many particles can exhibit phases or transport properties absent from a single-particle description.
• What can a measurement establish? Instrument response and uncertainty must be distinguished from the phenomenon being investigated.
• Where does a model cease to work? An approximation can fail at another size, speed, temperature, or level of precision.
• Which phenomena can be controlled usefully? Applied work connects physical mechanisms to measurable performance under stated constraints.

Instead of asking why oscillators behave strangely, a student could investigate how a pendulum’s period varies with its starting angle. Progress would mean a prediction, controlled measurements, and a defensible account of the conditions under which the model agrees with observation.

5. Core Concepts, Mental Models, and Working Assumptions

• Conservation and system boundaries: energy accounting asks where energy enters, leaves, or changes form. An LED converts electrical input into light and heat. Conservation does not mean its light output equals its electrical input.
• Fields and superposition: fields assign physical quantities throughout space and time. In linear optics, overlapping field amplitudes add, sometimes producing interference. Nonlinear material responses can invalidate that simple rule.
• Symmetry: a transformation leaves relevant laws unchanged. Noether’s 1918 theorem relates suitable continuous symmetries to conservation under specified mathematical conditions. Visual symmetry alone establishes no conservation law.
• State and evolution: a model specifies starting information and rules for change. Classical positions and velocities differ from a quantum state, which supports predictions of measurement probabilities. Quantum probabilities are not simply evidence of poor instrument calibration.
• Statistical ensembles and entropy: many microscopic arrangements can correspond to one macroscopic description. Schroeder’s An Introduction to Thermal Physics explains how counting such possibilities constrains thermal behavior; equilibrium formulas need not describe a driven system.
• Effective models and scale: a model deliberately retains relevant detail. Treating a pendulum as a point mass can be useful, but small-angle, negligible-friction, and rigid-support assumptions need separate checks.

These concepts help students ask not only whether an equation can be solved, but whether it represents the physical situation adequately.

6. Methods of Inquiry and Standards of Evidence

An investigation begins with a question, measurable quantities, a physical model, and an account of which observations would challenge it. Different approaches support different conclusions.

• Analytical theory: derive consequences from equations, symmetries, and approximations. A valid derivation establishes what follows from assumptions, not that nature necessarily satisfies them. Limiting cases and dimensional consistency provide important checks.
• Experiment: manipulate conditions, calibrate instruments, record controls, and repeat measurements. Taylor and Kuyatt’s NIST Technical Note 1297 (1994) distinguishes uncertainty evaluated statistically from uncertainty evaluated by other means. This distinction is not simply “random versus systematic”; calibration effects also carry uncertainty.
• Observation: investigate systems that cannot readily be manipulated, including astronomical sources. Spectra, images, and time series support inference, but instrument selection, incomplete coverage, and competing models constrain interpretation.
• Computation: approximate equations, simulate systems, or analyze measurements. Check numerical convergence, benchmark against solvable cases, and document algorithms. Agreement between two implementations is useful, but does not itself validate the underlying physical model.
• Applied development: build and test a device or measurement method against specifications. Gain, bandwidth, noise, thermal behavior, and robustness become evidence. Horowitz and Hill’s The Art of Electronics illustrates why real component behavior matters alongside ideal equations.

Working materials include notebooks, specimens, apparatus settings, calibration records, raw signals, code, and primary papers. Original measurements must be distinguished from reprocessed archives and simulated data. Strong work preserves that chain of evidence, explains exclusions and corrections, and supports independent checking rather than merely producing an attractive fitted curve.

7. Major Discoveries and Intellectual Contributions

• A unified electromagnetic account: Maxwell’s 1865 framework connected electricity, magnetism, and light. It made propagating fields central to understanding radiation and communication. Classical electromagnetism remains powerful, but does not by itself explain phenomena requiring quantized light.
• Microscopic explanations of thermal behavior: statistical mechanics connects particle states and probabilities to macroscopic quantities. Schroeder’s 2021 Oxford edition develops examples involving heat capacity and thermal machines. These explanations generate testable predictions, but equilibrium and interaction assumptions constrain their application.
• Symmetry as a testable physical claim: Chien-Shiung Wu, Ernest Ambler, Raymond Hayward, Dale Hoppes, and Ralph Hudson’s 1957 Experimental Test of Parity Conservation in Beta Decay reported directional asymmetry in electrons emitted by polarized cobalt-60. Weak beta decay did not preserve the assumed mirror symmetry. This did not invalidate Noether’s theorem: a conditional mathematical result does not guarantee its physical premises.
• Semiconductor amplification: John Bardeen and Walter Brattain’s 1949 Physical Principles Involved in Transistor Action described amplification in a germanium point-contact device. It exemplifies the meeting of condensed-matter explanation and device experimentation. Turning such behavior into dependable products also requires manufacturing control, circuit engineering, and reliability work.

8. Debates and Ethical Responsibilities

Prediction versus interpretation: the quantum formalism taught by Griffiths and Schroeter (2018) yields measurement probabilities, while interpretations can disagree about what the state represents. A proposed modification becomes empirically assessable when it predicts distinguishable outcomes. Explanatory disagreement does not justify ignoring successful calculations.

Detailed versus effective models: a microscopic simulation may represent mechanisms richly but contain poorly constrained parameters. A simpler model may predict reliably while concealing mechanisms. Thoughtful researchers can disagree because explanation, extrapolation, and control place different demands on a model.

Openness versus restricted use: sharing data and methods can conflict with intellectual-property, privacy, or security restrictions, especially for sensors and nuclear or defense technologies. Restrictions do not justify misleading reporting. The American Physical Society’s Ethics Standards calls for meaningful attribution, retained research records, correction of errors, and disclosure of relevant conflicts. Students should preserve processing choices, credit technicians and collaborators, and report unfavorable results. Equipment work also requires training, supervision, and hazard-appropriate controls; a successful calculation is not a safety approval.

9. How This Field and Adjacent Fields Investigate Shared Questions

The following comparisons are illustrative investigations, not reports of completed studies.

A candidate gravitational-wave signal

Suppose two instruments record similar brief disturbances. Physics and applied physics ask whether space-time dynamics, rather than an instrumental effect, explains the records. General relativity supplies candidate waveforms; applied physics examines interferometer response, calibration, and environmental coupling. Comparing synchronized records, monitoring channels, and predicted signals could yield a physically supported account with quantified limitations. A waveform match alone does not validate every assumption. The LIGO Scientific Collaboration and Virgo Collaboration’s 2016 Observation of Gravitational Waves from a Binary Black Hole Merger provides a documented methodological example.

Astronomy asks what source and formation history could produce the same candidate. It combines inferred source properties with astrophysical models and population evidence. This could place the event within stellar evolution, but uncertain histories and detector selection constrain claims about how frequently particular systems occur.

Statistics asks how strongly the records distinguish a signal from a specified noise process. Likelihood comparisons, background estimates, and uncertainty intervals assess evidence. Results remain conditional on the models: an estimated false-alarm rate is not an unconditional probability that an explanation is true.

Instrument knowledge disciplines the noise model, statistical checks constrain interpretation, and astronomy supplies plausible source alternatives. The disciplines share methods without asking identical questions.

A temperature-sensitive photodetector

Consider a light detector whose voltage changes as it warms under constant nominal illumination. Physics and applied physics seek mechanisms: semiconductor response, thermally generated current, optical changes, or readout drift. A temperature-controlled investigation would separately measure source stability, detector output, and electronics behavior. Models could identify a plausible mechanism and operating range, but several mechanisms may fit limited data. Mark Fox’s Optical Properties of Solids (2010) supplies relevant light–matter concepts; Horowitz and Hill’s The Art of Electronics (2015) addresses nonideal readout circuits.

Electrical engineering examines the same detector within a circuit and system. Feedback, gain, bandwidth, noise, and power constraints guide controlled circuit tests and redesign. The contribution could be a stable readout architecture; compensating the output does not necessarily identify the underlying material mechanism.

Materials science examines composition, defects, interfaces, processing, and structure–property relationships. Characterization of comparable samples and fabrication histories could connect drift to a material feature. Suitable instruments and controlled samples matter; an observed association alone would not establish the responsible microscopic process.

Together, these approaches distinguish explaining a material effect from compensating for it and from demonstrating performance in a complete device. That distinction helps students choose appropriate collaborators and evidence.

10. What Students Learn and Where the Field Can Lead

U.S. four-year study usually progresses from calculus-based mechanics, electromagnetism, and introductory laboratories to advanced classical physics, quantum mechanics, thermodynamics, computation, and experimental work. Calculus, linear algebra, and differential equations supply essential tools.

Program differences matter. Yale’s 2026–27 catalog requires research in both its Physics B.S. and Applied Physics B.S.; the latter combines physics with engineering-oriented focus areas. Columbia’s 2026–27 Applied Physics B.S. includes at least two advanced laboratory courses. Northeastern’s 2026–27 catalog offers an interdisciplinary Applied Physics B.S. in Boston with a senior capstone. These are examples, not universal requirements.

Ask how much laboratory access students receive, who supervises substantial projects, whether code and uncertainty analysis are assessed, and how electives support deeper study. Physics and applied physics overlap extensively; compare actual curricula rather than assuming that one is purely theoretical and the other exclusively practical.

Complementary electronics, statistics, programming, materials chemistry, and writing can broaden preparation. Optical characterization connects with the Semiconductors & Semiconductor Equipment Practice; heat-transfer investigations connect with the HVACR Equipment (HVAC, Refrigeration) Practice. Seek supervised projects with inspectable designs, data, and tests.

Further directions include graduate research, instrumentation, scientific software, education, technology development, and technical analysis. Each requires role-specific preparation; coursework exposure is not demonstrated professional proficiency.

11. Examples of Substantial Senior-Year Assignments

These original, illustrative assignments are not claims about any university’s requirements. Each assumes appropriate instructor support and several weeks of sustained work.

Pendulum dynamics beyond the small-angle approximation

Context and task: In a senior computational mechanics course, determine how a pendulum’s period changes with initial angle and where the small-angle approximation misses a stated accuracy target. Develop a physical explanation rather than merely plotting numerical output.

Materials and approach: Use instructor-supplied parameters, the nonlinear equation of motion, and an approved numerical environment. Derive an energy-integral or perturbative benchmark, compare two integration schemes, check timestep convergence, and track energy in the undamped case. Add a separately labeled weak-damping extension after validating the baseline.

Scope and deliverables: Work individually for eight weeks over a bounded range of starting angles. Submit a reproducible code repository, automated checks, period-versus-angle plots, and an eight-to-ten-page report defining the accuracy criterion and discussing assumptions.

Assessment: Evaluate mathematical formulation, separation of physical and numerical error, convergence evidence, readable code, and honest limitations. A polished curve without a validated calculation is insufficient.

Measuring a resonant circuit

Context and task: In an advanced experimental physics laboratory, investigate how a low-voltage resistor–inductor–capacitor circuit responds to a changing driving frequency. Determine whether an ideal resonance model adequately explains measured amplitude and phase.

Materials and approach: The department supplies approved components, a signal generator, oscilloscope, multimeter, instrument manuals, and laboratory training. After an instructor reviews the circuit, measure several resistive configurations. Compare ideal predictions with models incorporating component resistance and instrument loading. Preserve raw readings and propagate relevant uncertainties.

Scope and deliverables: Work in pairs for six weeks, sharing measurements but writing individual interpretations. Submit laboratory notebooks, raw data, fitting code, an uncertainty budget, a seven-to-nine-page paper, and a ten-minute presentation explaining model choice and any unresolved discrepancies.

Assessment: Judge calibration, experimental control, physical reasoning, repeatability, and transparent reporting. Agreement with an expected resonance curve is not the grading objective; a well-supported account of disagreement can be equally strong.

A calibrated optical sensor prototype

Context and task: In a senior applied physics instrumentation course, build a photodiode-based light sensor and evaluate a specified trade-off among sensitivity, bandwidth, and noise. The instructor supplies performance targets, not an assumed successful outcome.

Materials and approach: Use department-supplied photodiodes, low-voltage electronics, an LED illumination fixture, approved test equipment, and a reference detector with stated calibration uncertainty. Compare two readout gains, investigate ambient-light leakage, and test performance at illumination settings withheld from calibration. No lasers, exposed mains wiring, or human-subject measurements are required.

Scope and deliverables: Teams of three work for ten weeks on one prototype. Submit the working device, schematic, bill of materials, data and code, a six-to-eight-page calibration report, an operating-limit sheet, and a twelve-minute demonstration.

Assessment: Evaluate justified design choices, safe execution, uncertainty analysis, reproducibility, and communication of limitations. Missing a performance target is acceptable when the diagnosis and evidence are rigorous.

12. Foundational Books and Other Essential Works

1. Richard P. Feynman, Robert B. Leighton, and Matthew Sands, The Feynman Lectures on Physics (three volumes, 1963–1965). A conceptually rich introduction connecting physical explanations and mathematical models. Begin with Volume I’s “Atoms in Motion” and “Conservation of Energy.” The lecture style can be demanding; use it alongside worked problems, not as a complete account of current research.

2. Albert Einstein, On the Electrodynamics of Moving Bodies (1905; originally published in German). An influential original argument for special relativity, organized around operational questions about clocks and motion. Start with the kinematical part, especially “Definition of Simultaneity.” Historical notation requires patience, and this paper is not a presentation of general relativity.

3. Emmy Noether, Invariante Variationsprobleme (1918); M. A. Tavel’s English translation, Invariant Variation Problems (1971). A foundational connection between continuous symmetry and conservation. Begin with “Preliminary Remarks and Formulation of Theorems.” This is advanced reading requiring calculus of variations; distinguish its mathematical conditions from claims that particular physical symmetries hold.

4. Barry N. Taylor and Chris E. Kuyatt, Guidelines for Evaluating and Expressing the Uncertainty of NIST Measurement Results, NIST Technical Note 1297, 1994 edition. A concise methods reference for reporting defensible measurements. Start with Type A and Type B uncertainty evaluations. It clarifies why uncertainty is not identical to error or repeated-measurement scatter.

5. David J. Griffiths and Darrell F. Schroeter, Introduction to Quantum Mechanics, third edition (Cambridge University Press, 2018). A structured route through wave functions, operators, and approximation methods. Begin with “The Wave Function,” then “Time-Independent Schrödinger Equation.” Calculus, complex numbers, and differential equations help; learning the formalism does not settle every interpretive question.

6. Daniel V. Schroeder, An Introduction to Thermal Physics (Oxford University Press reissue, 2021). Connects energy accounting, entropy, statistical reasoning, and thermal machines. Start with “Energy in Thermal Physics” and “The Second Law.” It provides a bridge between physics and engineering applications while making equilibrium assumptions important objects of scrutiny.

7. Mark Fox, Optical Properties of Solids, second edition (Oxford University Press, 2010). Connects optical measurements with electronic and structural models of materials. Begin with the introduction, especially optical coefficients and optical materials. Best approached after introductory electromagnetism and quantum physics; it is not a comprehensive guide to fabrication or device qualification.

8. Paul Horowitz and Winfield Hill, The Art of Electronics, third edition (Cambridge University Press, 2015). A practical account of nonideal components, feedback, noise, and readout design. Start with “Foundations,” then consult “Precision Circuits” alongside laboratory work. Understand the reasoning rather than copying circuits; component choices and safety decisions require appropriate current documentation.

13. How This Field Can Contribute to a Consulting Career

A physical-sciences background can help a consulting team distinguish a feasible mechanism from an unsupported performance claim. Its contribution depends on demonstrated work: coursework supplies models, not manufacturing, commercial, clinical, or project-delivery experience.

• Physical-limit estimates and experimental planning: these can support R&D portfolio prioritization. In the Renewable Power Generation (Solar, Wind, Hydro) Practice, an absorption or heat-flow model could inform a technical evidence map and staged test plan. Commercial attractiveness still requires market and financial analysis.
• Calibration and controlled comparison: these support Root cause analysis by distinguishing performance variation from measurement error and proposing discriminating tests. Possible outputs include a test matrix, competing causal explanations, and a documented recommendation developed with operating staff.
• Connecting mechanisms to prototypes: this can inform a Product development roadmap. Within the Medical Devices & Diagnostics Practice (Capital Equipment & Disposables/IVD), optical readout knowledge could shape performance milestones and validation questions. Clinical usefulness, regulatory strategy, usability, and production readiness require specialist contributions.
• Dynamical and thermal modeling: this can contribute to Reliability improvement. For the Space Systems, Launch & Satellite Operators Practice, outputs might include thermal sensitivity analysis and an environmental-test rationale. Mission assurance and lifetime claims require spacecraft expertise and qualified testing.
• Performance constraints and approximations: these can inform Design-to-value by comparing optical layouts, cooling options, or tolerances while preserving essential specifications. A trade-off matrix and uncertainty-aware cost model require manufacturing, procurement, and finance inputs beyond academic physics.

Preparation should add statistics, sound software practice, accounting, cost modeling, project coordination, sector knowledge, and concise writing. Section 11’s circuit and sensor assignments could demonstrate careful testing; the pendulum assignment could demonstrate model validation. None establishes competence to approve production devices, direct clinical services, or certify safety. Early delivery experience should involve experienced technical and consulting supervision.

Illustrative example: a factory’s light-sensor test station

A manufacturer cannot tell whether rejected units are defective or its test station is unstable. A physics-trained analyst, working with a quality engineer, could audit reference sources, temperature histories, calibration records, and raw detector signals. Controlled comparisons between stations could inform a measurement-system assessment and retest protocol. The decision is whether to change the test, redesign the device, or gather more evidence. Coursework alone does not establish production acceptance limits or demonstrate long-term reliability.

Illustrative example: choosing a photovoltaic research pilot

A developer is choosing between two early-stage photovoltaic approaches. An applied physicist could compare absorption, thermal behavior, and documented test conditions, then propose experiments with predeclared decision thresholds. The deliverable would be a staged pilot plan and assumption register feeding an investment case, not an assertion of bankable energy yield. Engineering teams would need to test outdoor performance and grid compatibility; financial specialists would assess costs and funding. A laboratory advantage may disappear at scale, and missing evidence could justify postponing the choice.

Physics contributes a distinctive discipline: make the mechanism explicit, identify its scale and assumptions, and demand measurements capable of changing the decision.