Lecture 6 — Genomics and Enhancement

Part B: Engineering Persons — CRISPR, Enhancement, and Identity

Author

Affiliation

Brendan Shea, PhD

An Introduction to Bioethics.

0.1 Would You Select?

You and your partner are doing IVF. Six embryos make it to the testing stage. The clinic offers a new add-on for $50,000. They will read each embryo’s whole genome and rank them on predicted IQ, height, and risk for common diseases. You can pick the embryo at the top of the list. The doctor stresses that it is legal, voluntary, and currently used by other couples in the clinic.

Write or discuss:

Is this medical treatment, disease prevention, or something else? Does the label matter?
The expected IQ gain from picking the top embryo is about 2 points. Does that change your answer?
If you would not pay $50,000 for this, but it became free tomorrow, would you take it then?

This scenario is not hypothetical. US fertility clinics began offering whole-embryo polygenic scoring around 2020, and the companies selling it appear later in this chapter.

0.2 After You’ve Discussed

Note

Treatment restores normal function — an antibiotic for an infection, insulin for diabetes.
Prevention keeps you out of harm — a vaccine, a statin.
Enhancement improves you beyond the normal range — sharper memory, taller body, longer life.

0.3 Today

By the end of Part B, you will be able to:

State three accounts of the treatment-enhancement distinction and explain why each one keeps collapsing.
Tell somatic from germline editing, and explain why this does most of the moral work in the gene-editing debate.
Describe what Casgevy (FDA 2023) does, and why somatic gene therapy is much less contested than germline editing.
Compare three positions on enhancement: Savulescu’s procreative beneficence, Sandel’s giftedness argument, and Habermas on making vs. growing.
Explain the non-identity problem and apply it to a real case.
Describe what polygenic embryo selection is doing in US fertility clinics today, and what it can and cannot deliver.

1 The Treatment-Enhancement Line

1.1 Why This Line Matters in Practice

The line between treatment and enhancement does work in three places.

Insurance. Health insurers pay for treatment. They do not pay for enhancement. Whether a procedure is one or the other decides who can afford it.
Professional duty. Doctors have an obligation to treat. They do not have an obligation to enhance. A patient cannot demand enhancement the way she can demand treatment.
Regulation. The FDA approves drugs as treatments for specific conditions. Off-label and enhancement uses fall outside its main review.

Move a procedure from one side of the line to the other, and three different systems respond.

1.2 A Working Definition

The standard definition in the philosophical literature is precise.

Enhancement = biomedical interventions used to improve human form or functioning beyond what is necessary to restore or sustain health.

A few implications travel with that definition.

It is the use, not the technology, that makes something an enhancement. The same drug can be treatment for one patient and enhancement for the next.
Most enhancement-capable drugs and procedures enter medicine first as treatments. They show up as enhancements later, off-label.
The definition rests on what counts as “health”, which is itself a matter of debate. (Moseley and Juengst 2025)

1.3 Three Accounts of the Line

Account	Treatment is…	Where it breaks
Professional-domain	Whatever the profession agrees to do	The profession’s scope grows with demand: shyness became social anxiety disorder, aging became “low T”
Normal-function (Boorse, Daniels)	Restoring species-typical function	Prevention: a vaccine pushes immunity above normal on purpose, yet is plainly medicine
Disease-based (Gert, Culver, Clouser)	Addressing a diagnosable malady	Nosological elasticity: disease categories expand to license the interventions people want

Each account does real work in clinical practice, and each has a failure it cannot fix (Moseley and Juengst 2025; Boorse 1977; Gert et al. 2006).

1.4 The Professional-Domain Account

The idea. Treatment is whatever doctors agree they should do. Enhancement is whatever they refuse. Medical societies set the line by professional consensus.

Why it collapses. Medicine has a track record of medicalization — absorbing new conditions into its scope as cultural demand grows. Shyness became social anxiety disorder. Restlessness in boys became ADHD. Aging in men became “low T.” Each was once outside medicine. Each is now inside.

If the profession’s own boundaries shift with demand, the profession’s word for what counts as treatment cannot be the bedrock of the distinction.

1.5 The Normal-Function Account

The idea (Boorse, Daniels)

Health is the ability to do what people of your age and sex can typically do. Treatment restores you to that range. Enhancement pushes you above it.

This account anchors the distinction in physiology, not professional opinion.

Where it collapses

Prevention. A vaccine pushes your immune response above normal on purpose. Statins lower cholesterol below the average for your age. Both look like enhancement under this account, but are clearly treatment in practice.
Disability. A powered wheelchair that travels faster than a healthy walker counts as enhancement here. That verdict seems wrong.

(Boorse 1977)

1.6 The Disease-Based Account

The idea (Gert, Culver, Clouser). Treatment addresses a malady — a condition that involves real harm and has no good external cause. Enhancement acts on a healthy system that has no malady to address.

Why it collapses. The list of recognized maladies keeps growing — and often grows to match what the profession (and the market) want to provide. This is nosological elasticity: coining new diseases to license new interventions.

For example, ADHD diagnosis rates have moved in lockstep with stimulant prescriptions, which have moved in lockstep with classroom and workplace demand for focus. Each step was inside the disease-based account. The aggregate looks more like enhancement.

(Gert et al. 2006)

1.7 Why the Line Keeps Collapsing

Each account does something the other two cannot.

The professional-domain account is honest about who actually decides. It gives no principled answer.
The normal-function account is honest about biology. It collapses on prevention.
The disease-based account is honest about clinical practice. It collapses under medicalization.

The distinction does not vanish, but it cannot be settled from the outside. (Moseley and Juengst 2025)

2 From Somatic to Germline

2.1 Somatic vs. Germline

Somatic changes

In one tissue, in one person.
Not in sperm or egg cells.
Not inherited.
The change dies with the patient.
All currently approved gene therapies are somatic.

Germline changes

In every cell of the resulting person, including reproductive cells.
Heritable. Passed to children.
Each descendant inherits it.
Banned for clinical use in nearly every country.
The He Jiankui case is the only known human germline edit to date.

The same molecular tool can do either. The moral weight is in which cells it touches.

2.2 CRISPR-Cas9 — How It Works

CRISPR-Cas9 is, in plain terms, a programmable find-and-replace for DNA. You design a short guide molecule that matches the sequence you want to change. The guide drags Cas9 (a cutting enzyme) to that spot. Cas9 cuts. The cell tries to repair the cut, and the edit ends up in place.

2.3 Casgevy (FDA, December 2023)

The first CRISPR therapy approved in the US treats sickle-cell disease, which affects about 100,000 Americans, most of them Black.
Doctors harvest the patient’s stem cells, edit them to switch on a different form of hemoglobin, and reinfuse them after weeks of chemotherapy.
It is a somatic edit: it changes the patient, not her future children.
The list price is about $2.2 million, and Medicaid coverage varies by state. The access question is a justice question.

2.4 Luxturna and the Broader Picture

Luxturna (FDA 2017) treats an inherited form of blindness caused by a mutation in the RPE65 gene. A working copy of the gene is injected directly into the retina. About $425,000 per eye.

Several other somatic gene therapies are now approved or in late-stage trials — for hemophilia, for spinal muscular atrophy, for some inherited immune disorders.

All of them are clearly treatment under all three accounts of the line. None of the moral arguments later in this lecture apply to them.

The hard cases start when the same tools are used in embryos — or in healthy people.

2.5 He Jiankui — Briefly

What he did. In November 2018, Chinese scientist He Jiankui announced he had used CRISPR to edit the CCR5 gene in human embryos. He had implanted the embryos. Twin girls — Lulu and Nana — had been born. A third edited child was born shortly after.

His stated goal was to confer resistance to HIV. The edit was performed without serious clinical justification, with deeply flawed consent, and outside any sanctioned oversight.

He was sentenced to three years in prison in 2019 and released in 2022. The international consensus against clinical germline editing has held so far.

Paired case: He Jiankui’s CRISPR Twins — the full case study, including the consent failures and the scientific community’s response.

3 Cognitive Enhancement

3.1 Off-Label Stimulants in Healthy Adults

The most common enhancement today involves a prescription pad, not CRISPR. Roughly 1 in 6 US college students reports nonmedical use of stimulants like Adderall, with higher rates in nursing, medical, and law programs.
In diagnosed ADHD, stimulants clearly improve attention.
In healthy adults, the effect is small: a modest boost on simple attention tasks and almost none on complex problem-solving. The largest effect is on perceived productivity.
Side effects are real: insomnia, anxiety, raised heart rate, dependence.
US childhood ADHD diagnoses rose from about 3% in 1997 to about 11% in 2022, as the criteria broadened.

3.2 ADHD Diagnosis Over Time

Year	US children diagnosed with ADHD	Adult stimulant prescriptions
1997	~3%	uncommon
2007	~7%	rising
2016	~10%	common in adults
2022	~11%	sharply higher; telehealth-driven

The diagnosis rate has roughly tripled in twenty-five years. The criteria have broadened in each DSM revision since 1994.

Some of that increase is real recognition of a real condition. Some of it is the disease-based account collapsing in slow motion: as demand for the drug grew, the criteria for the diagnosis that justifies the drug expanded to match. Nosological elasticity in real time.

3.3 Thought Question — A Roommate’s NCLEX Request

Your roommate, a healthy nursing student with no ADHD diagnosis, asks to borrow one of your Adderall pills the night before NCLEX. She is anxious. She has heard from upperclassmen that “everyone uses it.” You are prescribed it for a diagnosis you have had since high school.

Write or discuss.

Apply the three accounts of the treatment-enhancement line to her request. Which one makes the strongest case for “no”?
Does the answer change if she says she just wants to try one pill at a normal dose?
If you would say no to your roommate, would you say no to yourself using the same drug to gain an edge on the same test?

3.4 After You’ve Discussed

Note

The equity angle. If passing NCLEX increasingly requires stimulants, the test starts measuring access to a diagnosis, not nursing knowledge. An enhancement that everyone uses stops being an advantage and becomes a requirement.

The three accounts give three different reasons for “no.” Professional-domain: prescribing is a professional act, and sharing a prescription with someone undiagnosed is outside it. Normal-function: the drug pushes attention above the normal range in a healthy user. Disease-based: she has no malady, so there is nothing to treat.

4 Who Is Wronged? The Enhancement Debate

4.1 Parfit’s 14-Year-Old Girl

Important

The thought experiment

Imagine a 14-year-old girl who is thinking about having a child. If she has the child now, the child will have a difficult start — limited resources, an inexperienced parent, predictable hardships. If she waits until she is 22, she will be more prepared, more stable, and the child she has then will have a much better life.

It seems obvious that waiting is the right thing to do.

But here is the puzzle. The child she would have at 14 is not the same person as the child she would have at 22. Different egg, different sperm, different person. If she waits, the first child never exists. He doesn’t get a better life; he simply does not get a life at all.

— Derek Parfit, Reasons and Persons (1984)

4.2 Parfit’s Puzzle

So who was wronged by the early pregnancy?

Not the actual child. His alternative was nonexistence. He is glad to be alive. He has no complaint.
Not the never-existing child. That child has no complaints, because that child never existed.

We feel a strong intuition that the early pregnancy was wrong. However, the standard person-affecting reasoning cannot say who was wronged. (Parfit 1984)

This matters everywhere we choose the conditions of someone’s existence — embryo selection, germline editing, even decisions about when to conceive.

4.3 He Jiankui’s Twins Through This Lens

Lulu and Nana exist only because their parents took part in He Jiankui’s experiment. A different embryo would have meant a different child — or no child at all.

So on the simplest reading: the twins were not wronged by being edited. The alternative was that they not exist.

That cannot be where the analysis ends. Most people share a strong sense that what He did was wrong. The non-identity problem forces us to locate the wrong somewhere other than “harm to the specific child who exists.”

Two candidate locations show up later in the lecture: a wrong to the kind of relationship parents and children should have (Sandel, Habermas), and a wrong to the future adult who was set up without consent (Feinberg, Habermas).

4.4 The Central Question

If safe, accurate gene editing or embryo selection were available — and many forms already are — should parents use them to give their children advantages they would not otherwise have?

Three answers from the philosophical literature:

Savulescu: yes, and parents have a positive duty to do so.
Sandel: no, because doing so corrodes something valuable in the parent-child relationship.
Habermas: no, because the child cannot consent to being someone else’s project.

4.5 Savulescu — Procreative Beneficence

The argument. If you are going to bring a child into existence, you have a moral reason to bring the child with the best expected life. The reasoning is straightforward consequentialism: more well-being is better than less well-being, all else equal.

The argument applies the same way to disease risk and to traits. If a parent should pick the embryo less likely to develop cancer, the parent should also pick the embryo more likely to be tall, intelligent, or cheerful — to the extent those traits track well-being.

Savulescu calls this procreative beneficence. It is the most permissive position in the contemporary literature.

(Savulescu 2001)

4.6 Sandel — Verbatim

“To appreciate children as gifts is to accept them as they come, not as objects of our design, or products of our will, or instruments of our ambition. Parental love is not contingent on the talents and attributes the child happens to have.”

— Michael Sandel, The Case Against Perfection: Ethics in the Age of Genetic Engineering (2007), p. 45.

(Sandel 2007)

4.7 Sandel’s Giftedness Argument

The argument is not that enhancement makes children worse off. It often does not. The argument is about what enhancement does to the parent and to the practice of parenting.

A parent who accepts the child as a gift can love the child unconditionally. Effort goes into the child’s flourishing on the child’s own terms.
A parent who designs the child treats the child as a project. Love starts to track how well the project is going.
Beyond the family, a society of designed children loses what Sandel calls solidarity — the sense that we are all subject to the same lottery and owe each other something for it.

The argument is not against medicine. It is against the idea that parenting is a manufacturing problem.

4.8 Habermas — Making vs. Growing

For Habermas, the wrong of enhancement is not in the child’s outcomes. It is in the kind of being a designed child is.

A child who grew — under nature’s contingency — can come to understand her life as her own. The traits she has are not someone else’s choice.
A child who was made — selected or edited toward a specific design — is, in a real sense, the project of her parents’ lives. She cannot retroactively consent to that. She cannot un-design herself.

This is not a religious argument. It is instead an argument about the structure of autonomy. The future adult is owed a life that is hers, not an extension of someone else’s plan. (Habermas 2003)

4.9 Three Positions at a Glance

	Savulescu	Sandel	Habermas
Enhancement is	A duty, where it improves the child’s life	A corrosion of the parent-child relationship	A wrong to the future autonomous adult
What’s lost without it	Well-being the child could have had	Nothing — the giftedness is the point	Nothing — the autonomy is the point
What’s lost with it	Nothing intrinsic; only access concerns	Unconditional love; social solidarity	The child’s authorship of her own life
Polygenic IQ scoring is	Permissible; perhaps obligatory if affordable	The textbook example of the drive to mastery	The clearest case of making a person

All three can agree that Casgevy is fine, but split on embryo selection.

4.10 Open Future, Now for the Body

Part A used Feinberg’s open future for genetic information: don’t tell an eight-year-old her BRCA risk before she can decide whether to know.
The engineering version: don’t build her toward a specific future before she can decide what future she wants (Feinberg 1980).
This answers the non-identity puzzle for He Jiankui’s twins. They cannot say they were harmed, since the alternative was not existing. The wrong lies in the relationship the experiment set up with two future adults who could never agree to be designed.

4.11 The Future Person Who Never Consented

Habermas and Feinberg meet at the He Jiankui twins.

Habermas: they were made, in a way that makes their lives the project of someone else’s plan.
Feinberg: a future they might have chosen for themselves was foreclosed in the embryo dish.
Non-identity problem: they cannot complain that they were harmed, because the alternative was nonexistence.

The wrong, then, is not to the twins-vs-no-twins comparison. It is to the kind of relationship He Jiankui’s experiment set up between him, the parents, and two future adults who could never have agreed to be designed this way.

5 Polygenic Embryo Selection

5.1 What Polygenic Embryo Selection Is

No editing and no CRISPR, only selection among embryos that already exist.
Clinics have screened out single-gene diseases (Tay-Sachs, cystic fibrosis) for decades. What is new is selecting on risk of common disease, and on traits like height or predicted IQ.
Orchid Health sells whole-genome embryo screening across hundreds of conditions. Heliospect Genomics explicitly scores for IQ. Both operate legally in the US; no federal law forbids it.

5.2 From Disease Screening to Trait Selection

The standard preimplantation test screens for things almost no one wants — a fatal childhood disease, a chromosomal abnormality. The new market screens for things that look more like preferences.

Risk of adult heart disease.
Risk of type 2 diabetes.
Predicted height.
Predicted educational attainment.
Predicted IQ.

Each step looks like a small extension of the last. However, the aggregate has crossed a line the field used to police more carefully. There is no US federal law against it.

5.3 Orchid Health and Heliospect

Two US companies are the live edge of the field.

Orchid Health (San Francisco) offers what it calls “whole-genome embryo screening” — sequencing every embryo and reporting risk scores across hundreds of conditions, including some adult-onset ones. Prices range from a few thousand dollars per embryo upward.
Heliospect Genomics offers polygenic embryo selection that explicitly includes scoring for IQ. Reporting by The Guardian in 2024 documented the service in detail.

Both operate legally in the US. The market is small but growing. Neither product is available in most European countries.

5.4 How Accurate Is a Polygenic IQ Score?

Current polygenic scores for IQ explain about 10–15% of the variance in intelligence-test performance. The rest is environment, gene-environment interaction, and noise.
For a couple selecting among 6–10 sibling embryos, the expected IQ gain from picking the top-ranked one is roughly 2–4 points. Within the noise of the test itself.
Predictions work best in the population the score was trained on — overwhelmingly people of European ancestry, as Part A noted.
Almost everything that makes a person’s life go well — temperament, motivation, the household they grow up in — is not in the score.

The product is a small statistical nudge, not a designer baby.

5.5 Back to the Clinic

Return to the six embryos from the start of class. The top embryo’s predicted IQ is 2.3 points above the bottom one, and the counselor says “most of our families choose the top.”

Has your answer changed, now that you know what the score can and cannot do?
Suppose the gap were 10 points instead of 2.3. Does that change it?
If you would not select for IQ, would you select for low risk of type 2 diabetes? What is different about the two choices?

5.6 After You’ve Discussed

Note

What is actually selected. The “best” embryo is best only in expectation, and the expectation is small. The choice is a probability bet, not a guarantee.
What is left out. Most of what shapes a child’s life is environment and chance — neither of which is in the score. Selection promises agency over a slice of the outcome and quiet about the rest.
Equity. The service is currently sold at price points that exclude most American families. If it works at all, it works to widen advantage that money already buys.

5.7 Treatment or Enhancement?

Where it shows up	Looks like treatment because	Looks like enhancement because
Casgevy for sickle cell	Treats a clear malady; restores normal hemoglobin	Permanent edit; price puts it out of reach for most patients
Adderall off-label	Same molecule as the prescription drug	Acts on a healthy brain; gives a competitive edge
He Jiankui’s CCR5 edit	Framed as HIV prevention	Healthy embryos; no clear medical need; heritable
Polygenic disease screening	Selects against future illness	Treats embryos as products to be ranked
Polygenic IQ selection	Sold as “best chance for your child”	No malady; pure trait selection
Future germline editing	Could prevent inherited disease	Could redesign human capacities

5.8 Review Questions

Write a brief answer to each, or use them as discussion prompts.

Recall

State the three accounts of the treatment-enhancement distinction in your own words. For each, name one concrete example where it gives an unhelpful answer.

Apply

A young woman with sickle-cell disease is considering Casgevy, but her state’s Medicaid program may not cover it. Name two things she needs to understand about the treatment, and one ethical issue that goes beyond her individual case.

Debate

A US state legislature is considering a bill that would ban polygenic embryo selection for non-disease traits in fertility clinics. Take a position. Identify the strongest argument against your position and respond to it.

5.9 Key Terms

Treatment vs. enhancement — treatment restores normal functioning; enhancement improves a person beyond the normal range. The line matters for insurance, professional duty, and regulation — but all three standard accounts of where to draw it keep collapsing.
Normal-function account — Boorse and Daniels’s view that health is the ability to perform what people of your age and sex can typically perform; treatment restores that capacity, enhancement exceeds it.
Nosological elasticity — the tendency of disease categories to expand to match demand for the drugs or procedures that treat them, as seen in ADHD diagnosis rates tracking stimulant prescriptions.
Somatic vs. germline — somatic edits change one tissue in one person and are not heritable; germline edits change reproductive cells and pass to every descendant. The same molecular tool can do either; the moral weight is in which cells it touches.
CRISPR-Cas9 — a programmable gene-editing system that uses a short guide RNA to direct a cutting enzyme (Cas9) to a specific DNA sequence, where it makes a break the cell then repairs, leaving the intended edit in place.

Procreative beneficence — Savulescu’s principle that parents have a moral reason to select or engineer the child with the best expected life, applying consequentialist reasoning to both disease risk and ordinary traits.
Giftedness (Sandel) — Sandel’s term for the parental disposition to accept a child as she comes, rather than as the product of design; he argues that engineering children corrodes unconditional love and social solidarity.
Making vs. growing (Habermas) — Habermas’s distinction between a child who comes into existence through nature’s contingency (and can own her own life) and one who is designed toward a parental plan (and cannot retroactively consent to being someone else’s project).
Non-identity problem — Parfit’s puzzle: a person who exists because of a decision that seemed harmful cannot claim to be worse off than she would otherwise have been, because the alternative was that she would never exist at all.
Polygenic embryo selection — an IVF add-on that sequences multiple embryos, calculates polygenic risk scores for disease or traits, and lets parents choose which embryo to implant; no editing is performed — only selection among what already exists.

5.10 Further Reading

Michael Sandel, The Case Against Perfection: Ethics in the Age of Genetic Engineering (Harvard, 2007).
Jürgen Habermas, The Future of Human Nature (Polity, 2003).
Daniel Moseley and Eric Juengst, “Human Enhancement,” Stanford Encyclopedia of Philosophy (Winter 2025).
Julian Savulescu, “Procreative Beneficence: Why We Should Select the Best Children,” Bioethics 15 (2001).
James Tabery, Tyranny of the Gene (Knopf, 2023).

5.11 References

Boorse, Christopher. 1977. “Health as a Theoretical Concept.” Philosophy of Science 44 (4): 542–73. https://doi.org/10.1086/288768.

Feinberg, Joel. 1980. “The Child’s Right to an Open Future.” In Whose Child? Children’s Rights, Parental Authority, and State Power, edited by William Aiken and Hugh LaFollette. Littlefield, Adams.

Gert, Bernard, Charles M. Culver, and K. Danner Clouser. 2006. Bioethics: A Systematic Approach. 2nd ed. Oxford University Press.

Habermas, Jürgen. 2003. The Future of Human Nature. Polity Press.

Moseley, Daniel, and Eric Juengst. 2025. Human Enhancement. The Stanford Encyclopedia of Philosophy (Winter 2025 Edition), Edward N. Zalta and Uri Nodelman (eds.). https://plato.stanford.edu/archives/win2025/entries/enhancement/.

Parfit, Derek. 1984. Reasons and Persons. Oxford University Press.

Sandel, Michael J. 2007. The Case Against Perfection: Ethics in the Age of Genetic Engineering. Belknap Press of Harvard University Press.

Savulescu, Julian. 2001. “Procreative Beneficence: Why We Should Select the Best Children.” Bioethics 15 (5–6): 413–26. https://doi.org/10.1111/1467-8519.00251.