* Knight v. Knight (1840); + Lord Langdale - To be valid as a trust there must be:
a) Certainty of intention to create a trust – Is a trust intended as a q. of fact?
b) Certainty of subject-matter – What property is to be subject to the trust and what are the beneficial interests?
c) Certainty of objects – Who are the beneficiaries of the trust? (charitable trusts do not need to satisfy this requirement)
* Why are these requirements necessary?
* The essential elements of the trust relationship must be defined with sufficient certainty to enable the T or donee of the power, or by default the court, to carry out his/her duties. Ts must know what their obligations are under the trust. The settlor cannot put the T under a duty to do something that is too vague to be legally enforceable or that is practically unworkable, given the nature of the trustee’s duties and their liability for breach. Certainties provide T with a degree of protection.
* The certainty rules have to strike a balance: the settlor’s freedom to express his or her intentions about how the trust is to work must be set against the need for the trust to be workable as a trust.
NB. A trust still has to comply with the extra formality requirements to be valid.
1. Certainty of Intention to Create a Trust
Test: In all cases, it is a question of proving that the settlor intended to impose a legally binding obligation on the trustee to hold and manage the property on behalf of the B.
Equity looks to intent, rather than form. This is a question of construction of the relevant documents or of gathering inferences from the words or conduct of the alleged settlor, considering all the circumstances of the case.
Administration of Justice Act 1982, s.21: Extrinsic evidence, including evidence of the T’s intention, may be admitted to assist in the interpretation of a will.
Use of the word ‘trust’
* Kinloch v. S of S for India; There is no magic in the use of the word ‘trust’.
* Re Harrison (2005); In relation to a will Hart J said, that the mere facts that the words ‘in trust’ had been used was not, in itself, inconsistent with an intention that the testator’s wife should be the absolute beneficial owner. A strong context is required to deny the prima facie construction of the word ‘trust’.
* Re Harding ; Lewison J noted that Hart J decided that the words ‘in trust’ in the will that he was considering were incompatible with an absolute gift.
Imperative or precatory language? Since Lambe v. Eames (1871), the courts have generally made a distinction:
* Imperative words express a command, a duty to do something. Use of such words indicates that a trust (or power) is intended. * Precatory words express a hope, a wish, or a moral obligation that the donee will deal with testator’s property in a particular way. Use of such words typically indicates that a gift is intended. There is no intention to impose a binding trust on that person.
Modern attitude - No trust is created by precatory words:
* Lambe v. Eames (1871); ‘to be at her disposal in any way she may think best, for the benefit of herself and her family’. * Re Adams (1884); Held there was no trust created by a testator who gave all of his property to his wife ‘in full confidence that she would do what is right as to the disposal thereof between my children, either in her lifetime, or by will after her death’. + Cotton LJ: ‘Undoubtedly confidence…may make a trust, but what we have to look at is the whole of the will which we have to construe’.
* Re Diggles (1888); ‘it’s my desire that she allows A an annuity of £25 during her life’.
* Comiskey v. Bowring-Hanbury ; The presence of precatory words will not necessarily prevent the court from finding that a trust exists, as long as it is satisfied that this was the donor’s intention. Here, reading the ‘full confidence’ section as precatory, later words suggested H intended to convey a life interest to W. * Staden v. Jones ; Ct may also take into account surrounding evidence which sheds light on the parties’ intentions.
Here the CA looked to a solicitor’s covering letter to conclude that a divorcing couple’s arrangement that the W transferred her share in the family home to the H on the basis that their daughter should ultimately be entitled to her share amounted to a constructive trust.
* Paul v. Constance ; An intention to create a trust can also be inferred from the donor’s conduct. Facts: Dispute over whether Mr C’s wife (from whom he was separated but not divorced) or his new partner, Mrs P, was entitled to money held in a bank account in the deceased’s sole name. During their relationship Mr C had made arrangements for Mrs P to withdraw money with his permission.
Only Mr C withdrew money once, which was split evenly between them and he often told Mrs P that the money was ‘as much yours as mine’. They also paid some joint bingo winnings into the account. Decision: Held these actions were sufficient to infer that Mr C had made a declaration of trust of the money in the account and Mrs C was entitled to half of the account.
* Re Kayford Ltd (in liquidation) ; The separation of customers’ money, paid for the future supply of goods, into a different bank account was deemed sufficient to demonstrate an intention to create a trust pending delivery.
The effect of lack of certainty of intention
* Lassence v. Tierney (1849); Donee will take property absolutely as a gift if no intention.
2. Certainty of Subject-Matter
There are two elements to certainty of subject-matter: 1) It must be clear what property is held on trust; and 2) The beneficial interests of the cestuis que trust must be certain and clear.
Issue: + G. Williams (1940): Requirement of certainty of subject-matter is somewhat ambiguous.
(i) Identifying the asset which is the subject-matter of the trust –
Rule: The donor must make clear what property is to be held on trust or the trust will be void as the Ts, Bs and the court will be unable to know what is held on trust (Westdeutsche Landesbank , per Lord B-W). There can be a trust of almost anything: chattels, a chose in action, a right or obligation under a contract, land or money.
(a) The asset must exist -
The asset which is to be the subject matter of the trust must exist. The trust must fail if the defendant fails to constitute the fund to which the trust was supposed to attach.
* Fortex Group v. Macintosh  3 NZLR 171; Facts: Employees paid monthly into a pension fund. Money was supposed to be paid into a separate company account but this did not happen. Employer (F) became insolvent and had been using the sums paid into the pension fund in its general expenses. Issue: Had F declared a trust of the unpaid contributions?
Are the Bs secure creditors? Decision: Held no trust because, 1) there was never any intention on the part of F to declare a trust, they just breached their obligations; and 2) there was no identifiable subject matter that the trust could have bitten upon had it been declared. A trust can’t attach to a mere net surplus of monetary wealth, only to an identifiable fund of property. ----------------------------------------------------------------------------------------------------------------------------------------------
(b) Identifying the ‘part’ of the asset which is the subject-matter of the trust –
Situation: Where there is a trust over part of a bulk of tangible property, the trust property will only be certain if it has been separated from the rest.
Issue: In such a case, the part must be defined with sufficient certainty to distinguish it from the rest, which may belong beneficially to the trustee or to another beneficiary. How is the part sufficiently certain?
Cases where the subject-matter was too uncertain: Trust is void…
* Palmer v. Simmonds (1854); Facts: The trust was only to attach the ‘bulk’ of the testatrix’s estate. Decision: Trust altogether void because the definition of the subject matter was not clear enough. The proportions were not easily ascertainable and so the trust was practically incapable of being administered – no one could dispose of property as they would not know if it was their own or the trust’s. H would have no way of knowing in advance if he was committing a breach of trust as the extent of the trust fund is not defined clearly enough. Criticism: Arguably this is too technical, although Fox says it is not.
* Peck v. Halsey (1720); ‘some of my best linen’. * Re Kolb’s WT ; An instruction to purchase ‘blue chip’ investments.
Court attempts to make trusts valid:
* Re Golay’s Will Trusts ; Facts: Ct bent over backwards to make this trust of income work. Testator stated that his ‘friend’ would be allowed to take a ‘reasonable income’ from the trust; no specific quantity. Decision: Held to be valid and not too vague. Court said that it is perfectly familiar with what ‘reasonable’ means from conducting damages cases. Objective standard so trust for an undefined, but ‘reasonable’, income is calculable. NB. Another factual situation: If the testator says that the trustee should pay the friend whatever sum they think reasonable, that is valid too as it is a kind of discretionary trust.
What happens if the trust is void for lack of certainty of subject-matter?
* Hancock v. Watson ; If the subject-matter is held to be uncertain the residuary B, who is meant to take the absolute interest, takes it all entirely. In other words, if the portion meant for another cannot be identified, the trust will fail, but the residuary amount will pass to the main B. ----------------------------------------------------------------------------------------------------------------------------------------------
(c) A trust attaching to an unascertained number of units in a larger fund –
Tangible property S.20A Sale of Goods Act 1979; offers protection to purchasers of an unsegregated part of a larger bulk of property. As soon as they have paid, they will be considered tenants in common of the whole property. This has the effect of protecting their purchasers from creditors in the event of liquidation or receivership.
* Re London Wine Co. Ltd ; Facts: LWC stocked cases of wine in various warehouses. When wine was purchased by customers, they received a certificate of title; indicated the wine would be held on trust for them until dispatched. Intention was that the wine purchased should be the customers’ property, but stored by the co. at customers’ expense. LWC went into liquidation. Issue: Was the wine was held on trust for the customers or was to be considered part of LWC’s general assets available to creditors?
Decision: Held there was perhaps an intention to create a trust, but actually no trust, as the wine ordered by customers had not been separated from the general stock and so the subject-matter of each trust could not be identified. There was no appropriation of the wine within the warehouse until actual delivery to the purchaser. Justification: All the bottles of wine were potentially different among themselves. Principle: When property is tangible the trust property must be specifically identified. + Pettit: right in principle.
Vs. Intangible property However, where there is a trust of some intangible property, such as shares, there is no need to identify the specific shares to be held on trust.
* Hunter v. Moss (1994) (CA); Facts: M owned 950 shares in a co. He declared himself T of 50 of those shares for H. Later M sold all his shares and kept the proceeds. H sought a share of the proceeds. M argued that the trust was void because he had not separated or identified the specific shares to be held. Test: Held, in such cases, that the question of certainty depends not on the application of any immutable principle based on the requirements of the needs for segregation or appropriation, but rather on whether, immediately after the purported declaration of trust, the court could, if asked, make an order for execution of the purported trust. Decision:
Trust was workable. Dillon J distinguished Re London Wine Co. on the basis that, unlike cases of wine or other tangible property, these shares were indistinguishable from one another. They all carried similar rights to vote at company’s meetings etc. Hence no separation was required as holding any 50 shares would achieve the same thing.
Criticism: + Hayton: 1. Is Dillon J’s argument that there is no difference between a testator giving 50 shares to a legatee in his will and a settlor declaring himself T of his 50 shares correct? 2. Difficult questions may arise in the event that the ‘trustee’ sells part of his holding – E.g. Has he sold his own shares or the Bs or, if the transaction is taxable, who is liable, the seller or the B?
+ Worthington: Suggests such issues could be resolved through tracing. 3. Is the broad distinction in the cases between trusts of part of unascertained bulks of tangible and intangible property convincing? + Ockleton  CLJ 448:
+ Alison Jones Conv 466: Argues that intangible assets are not in a different position from tangible. Nor is the analogy drawn by the court with a demonstrative legacy valid, because a trust of such shares will become completely constituted only when the particular shares have been vested in the trustee. - J Martin  Conv 223:
- Worthington  JBL 1: Supports. Australian cases have declined to follow Hunter v. Moss: White v. Shorthall (2006).
* Re Goldcorp Exchange Ltd. ; Distinguished from Hunter also. This was a case of a trust over tangible property where some of the units might have varied between themselves.
Ways of getting out of the problem:
* Re Hallett’s Estate (1880); Artificial identification - presumption of identification is used by the court to untangle mass of property (tracing). Court would simply deem after the event that if the trustee sold 900/950 shares, and when he becomes bankrupt that there are 50 left, the court will simply deem that the 50 belongs to B.
* White v. Shortall (2006) (SCNSW); Can you make the trust work consistently with the testator’s wishes? Ct implied a duty to make the trust work - read in that trustee owns an implicit duty always to maintain a sufficient balance within the total fund to meet the amount of the B’s claim, if and when the B wants to enforce it. So the trust works by implying.
Is there a valid distinction in the standards of certainty between testamentary trusts and inter vivos trusts?
* Re Clifford ; Diff gift – a testamentary trust. Facts: Testator had provided the B of the will would get 20/104 shares in his company free. In Hunter the court says this is an example valid just like Hunter. Criticism: + Fox: Says it is not the same: an executor of a will, then one of your duties as an executor is to identify the property in T’s estate and then apply it towards the B’s gift. There is an inbuilt duty which makes the trust workable, even though the testator has not specifically identified the property. So unless there is a duty in a case like Hunter there is little chance the trust is workable.
Reality: Trusts like Hunter v Moss will not be attacked for being invalid as they are much too common and millions of pounds of funds depend upon their validity.
The implications of Hunter v. Moss  1 W.L.R. 452 for the ownership of unallocated holdings of shares in dematerialized share trading systems. Re Harvard Securities Ltd (in liq.); Holland v. Newbury  2 B.C.L.C. 369.
Is it enough that the trustee’s obligation to identify and segregate the units from the larger fund can be defined with sufficient certainty even though the settlor has not specifically identified the units to which the trust should attach?
______________________________________________________________________________________________ (ii) The beneficial interest which the B is to take must be clear – Is the beneficiary to take a simple life interest under the trust or a life interest coupled with a power to dispose of capital during his or her lifetime?
* E.g. T devises his bungalow to H and gives her a legacy or £1500 on terms that in her will she would leave the bungalow and any remaining money from the legacy to T’s son, O.
Problems that may arise in this context:
* Boyce v. Boyce (1849); Facts: Testator established a trust of 4 houses for his daughters, M and C. M was to choose which house she wanted and the other houses would be held on trust for C. M died before making her choice. Decision: Trust failed as it was no longer possible to say which houses would be held on trust for C. Property had been clearly identified but the beneficial interest had not. Trust would have succeeded had Ts been given the power to choose the house. No-one could determine.
NB. Legal issues surrounding the certainty of beneficial interests will not apply in respect of discretionary trusts. This is because the class to benefit merely hold a spes (hope) of benefiting – the extent to which they can benefit, if at all, at the absolute discretion of the Ts.
NB. If there is a lack of certainty as to the subject-matter, this will cast doubt on whether the settlor truly intended to create a trust (Mussoorie Bank Ltd v. Raynor (1882)).
* Ottaway v. Norman  Ch. 698; * Birmingham v. Renfrew (1936) 57 C.L.R. 666;
See Underhill & Hayton, Law of Trusts and Trustees (16th ed.) at p. 79. ______________________________________________________________________________________________
3. Certainty of Objects
The fundamental test: Can the trustee or donee of the power execute his or her duties under the settlement? Functional test – can you make the trust work? Can you make the trustee’s duties operable?
Rules: a) A power will be certain as to its objects if it can be said with certainty whether any given individual is or is not a member of the class. b) A fixed trust will be certain as to its objects if a complete list of Bs can be compiled. c) A discretionary trust will be certain as to its objects if it can be said with certainty that any given individual is or is not a member of the class.
Justification: Need for ascertainable beneficiaries.
What are the consequences of uncertainty of objects? If the trust fails then a resulting trust will probably come into existence so the beneficial right comes back to the settlor. Rule in Hancock and Wilson may apply.
a) Fiduciary Powers of Appointment – Duty to survey the group (unlike non-fiduciary powers)
* McPhail v. Doulton ; Lord Wilberforce: ‘The power is valid if it can be said with certainty whether any given individual is or is not a member of the class and does not fail simply because it is impossible to ascertain every member of the class’.
* Re Coates ; A will conferred a valid power on the wife – ‘I direct my executors to pay to such friend or friends as are nominated by my wife’.
* Re Baden’s Deed Trusts (No.2); Confused the issue. The majority held in effect that a power may be valid even though there may be a substantial no. of persons of whom it is impossible to say whether or not they are within the class or not. Three judges gave 3 opinions. Technically concerned discretionary trusts.
* Re Manisty’s Settlement ; Principle: Templeman J stated, ‘the mere width of a power cannot make it impossible for trustees to perform their duty nor prevent the court from determining whether the trustees are in breach’. However, a power (also assumed a discretionary trust) will fail if it is ‘capricious’.
‘Capricious’: Where the terms ‘negative any sensible intention on the part of the settlor’ such that the objects constitute, ‘an accidental conglomeration of persons who have no discernible link with the settlor or any institution’. The specific problem here is not to do with the breadth of the class definition or the no. of people who fall within it, but rather with who these people are and why the settlor picked them. The idea is that the settlor should have a good reason for choosing the objects, for if the class appears to be arbitrarily defined then the trustees will not be able to pick between them on any sensible basis when it comes to distribution.
This is a separate test from unworkability (District Auditor: trust failed for unworkability but held not to be capricious). It is the need for the trustees to be able to formulate sensible and coherent criteria for exercise of distributive discretion. ----------------------------------------------------------------------------------------------------------------------------------------------
b) Fixed trusts
Uncertainty in the definition of the words used to define the class of objects or in identifying which individuals should belong to it will cause the trust for fail for uncertainty of objects. Evidential difficulty will not cause the trust to fail – we only need to known the maximum number of Bs.
* IRC v. Broadway Cottages Trust ; A full list of Bs must be able to be created if the T is required to divide the fund into equal shares. All the Bs must be able to be identified because, if the total number is not known, the fund cannot be divided. Evidential difficulty will not cause the trust to fail – we only need to known the maximum number of Bs.
* E.g. S establishes a trust for L for life, remainder to her children in equal shares. Is ‘children’ sufficiently certain? It is necessary to prepare a complete list of all the children. The size of any one beneficiary’s share cannot be determined unless the total number of beneficiaries in the class is known.
* Re Benjamin ; The share of any beneficiary who can be identified but whose whereabouts cannot be ascertained should be paid into the court.
c) Discretionary trusts
* E.g. “The trustees shall apply the net income of the fund in making in their absolute discretion grants to or for the benefit of any of the officers and employees ... of the company or to any relatives or dependants of any such persons in such amounts …”
* McPhail v. Doulton (Re Baden’s Trust No.1) ; Facts: Settlor set up a fund for ‘the benefit of employees of his co. and their relatives and dependants’ at the ‘absolute discretion’ of the Ts. Issue: H of L had to decide: a) whether this was a trust or a power; and b) the appropriate test for the certainty of objects requirement. Decision: Held unanimously that this was a discretionary trust. 3:2 regarding the test: + Lord Wilberforce gave the leading judgment of the majority – 1. Held complete list approach of IRC v. Broadway Cottages Trust was overruled in the case of discretionary trusts.
2. Test for certainty of objects in discretionary trusts is same as test for fiduciary powers in Re Gulbenkian’s Settlement, namely whether it could be said with certainty that any given individual is or is not a member of the class. 3. Even where meaning of words used is clear, definition of Bs may be so hopelessly wide as not to form ‘anything like a class’, so that the trust is administratively unworkable (see below). Proposition does not apply to powers, where the courts can primarily determine whether a power is valid and, if so, whether an exercise is within its scope. Principle: Need conceptual certainty, but not evidential certainty.
Applying the test in McPhail v. Doulton:
* Re Baden’s Trust (No. 2) ; Issue: Whether the groups ‘relatives’ and ‘dependents’ satisfied the new test – i.e. could it be said with certainty that any given individual is or is not a member of these classes? Decision: CA held that only if a class of beneficiaries is conceptually certain will it be possible to satisfy the test of whether it can be said with certainty that an individual is in class.
NB. The same test of certainty of objects, as applied to fiduciary powers, also applies to discretionary trusts.
What is ‘conceptual certainty’? Has been described as ‘linguistic or semantic certainty’; must be able to define group clearly.
Gifts subject to condition precedent do not require the same degree of conceptual certainty.
Uncertainty in the general definition of the condition will not cause the gift to fail provided that the beneficiary can prove that, on any view, he or she satisfies the condition.
* Re Barlow’s WT ; Facts: Testatrix provided that a number of her paintings could be sold at a reduced price to ‘any members of my family and any friends of mine’. ‘Friends’ defines the class of objects. Issue: Whether the condition precedent (i.e. that they be family members or friends) rendered the gift void for uncertainty. Decision: Court upheld the gift. Whereas the objects of a discretionary trust must be conceptually certain, a condition precedent will be valid if at least one person can be said to satisfy the condition. Analysis: In a gift to a class, as the objects are only entitled to a share of the property, it is vital to be able to say whether any given individual is or is not a member of the class.
However, in this case, the ‘family and friends’ were each given an individual opportunity to purchase the paintings. This is not a problem here because you do not need to divide up the entire fund equally so you do not need to know the max no. of friends, or what ‘friend’ actually means. Therefore, a greater degree of uncertainty satisfied this condition and did not affect the opportunity they received.
= While a gift to ‘friends’ may be valid as a gift subject to a condition precedent, a discretionary trust for ‘friends’ will be void for uncertainty, as it is impossible to define friends so that it could be said that any individual is or is not within the class. Distinguishing conceptual uncertainty from evidential uncertainty
‘Conceptual uncertainty’ – Relates to the certainty of the class. Makes a trust void. ‘Evidential uncertainty’ – Relates to the issue of whether an individual can be found or proven to be a member of the class or not. Will not defeat a trust. * Re Baden’s Trust (No. 2); Issue: All judges agreed that ‘dependants’ was conceptually certain, and that ‘relatives’ was too but their reasoning concerning the latter differed:
+ Sachs LJ: Liberal approach - Began by defining ‘relatives’ to mean ‘descendants of a common ancestor’. Argued this was conceptually certain because it was clearly understandable. Stated Ts have a fiduciary obligation to survey the range of possible Bs. This involves the Ts gaining a sense of the general width of the class – it does not require an exhaustive list. Beyond that survey, anyone who can prove that they are a member of the class can be included.
Held that the McPhail test only required conceptual certainty; ‘the court is never defeated by evidential uncertainty’. Where the facts leave us unsure as to whether a person meets the definition, that person is to be treated as not falling within the class. ‘If it is not proved, then he is not in it’. ‘Don’t knows’ are ‘nos’. Allows Sachs LJ to say the McPhail trust is valid even if ‘relatives’ was to be understood as ‘descendants from a common ancestor’. The fact that it will not always be possible to prove that any given individual is not a member of a class will not render the trust void.
+ Megaw LJ: Middle ground - Argued that a class would be conceptually certain if it could be said with certainty that a substantial no. of objects fell within the class, even if there were a substantial no. of others of whom it could not definitely be said that they were within or without the class. Was willing to tolerate a class of ‘don’t knows’. He concluded that were it otherwise, and we had to answer ‘yes’ or ‘no’ to everyone then we would be back in the list system which had been clearly rejected in McPhail.
He did not agree with Sachs that evidential uncertainty was irrelevant though; there comes a point where, if the number of ‘don’t knows’ is substantial enough and evidential uncertainty is widespread, then the trust must fail. He upheld the trust. + Fox: Megaw’s approach is doubtful because the T still has a prior duty to survey the potential class of objections; they can’t just give money to the first person they find. So the ‘substantial no. of people’ does not fit with the nature of the survey.
+ Stamp LJ: Strict approach. Took the literal ‘is or is not’ McPhail approach. He regarded it as necessary that there be both a clear definition of class and sufficient factual info. A class of ‘don’t knows’ is not allowed. Stamp LJ declared the discretionary trust valid on the basis that the court could always determine who was a dependant and ‘relative’ because this could be legitimately restricted to ‘next of kin’.
+ McDonald: Like the list system. Undermines the liberal implications of McPhail.
= Still unresolved now. Read Emery (1982) 98 LQR 551.
Curing conceptual uncertainty:
Where a class as defined by the settlor is, or runs the risk of being held to be, conceptually uncertain, and hence fatal to the validity of the trust, it is not uncommon for the settlor to attempt to rescue the trust from uncertainty and invalidity by stipulating that uncertainties are to be resolved conclusively by a named third party.
* Re Tuck’s Settlement Trusts ; Facts: Settlor created a trust in favour of future baronets, provided they be married to a wife who is ‘of Jewish blood’ and who ‘continues to worship according to the Jewish faith.’ Settlor further provided that, where this was in doubt, ‘the decision of the Chief Rabbi in London either of the Portuguese or Anglo German community…shall be conclusive’. Issue: Was the trust void for uncertainty because of the inherent ambiguity in the references to Jewish blood and the Jewish faith? Decision: Lord Denning MR held that the Chief Rabbi provision was effective to resolve a conceptual uncertainty in these terms.
Eveleigh LJ’s reasoning rests on the narrower ground that there was no conceptual uncertainty because the settlor was ‘in effect saying that his definition of “Jewish faith” is the same as the Chief Rabbi’s definition’ (conceptually certain). Issue: What if that question is batted off to a third party and they come up with an untenable answer? There are grounds for review. Lord Denning mentions these in Re Tuck – so long as the determination made by the third party is not ‘wholly unreasonable’ and it is ‘made in good faith’ – that is ok.
Problems: * Other judges upheld the trust on other grounds so Denning’s opinion was minority. * Case addresses the validity of conditions precedent. Remains unclear whether the same approach would be adopted in respect of a discretionary trust with conceptually uncertain objects, or whether the courts would adopt the broad approach of Denning LJ or the more restrictive approach of Eveleigh J.
Although Denning’s would ensure the validity of many more discretionary trusts, it would also seem to empower individuals named by the settlor to cure conceptual uncertainty in cases where the ct would otherwise declare void. + Webb: Eveleigh’s approach maintains the need for conceptual certainty but rests on a fine distinction that many settlors will not appreciate. !! There is also a middle ground which gathers support from some cases (Dundee General Hospitals Board of Management v Walker  and Re Tuck, Lord Eveleigh LJ:
* Dundee General Hospital Board of Management v. Walker ; H of L held that if there is a mixed question of fact and law, even this question can be delegated to a 3rd party to get a conclusive determination.
Says that there are two different ways in which a third party could be called upon to resolve uncertainty in the definition of class: 1. Could be called upon to resolve uncertainty in the definition of the class (e.g. ‘£100 to my friends and, where there is any uncertainty, matters are to be resolved by X’). Here the class (‘friends’) is conceptually uncertain and the third party cannot save it. 2.
The third party’s opinion may itself form part of the definition of class provided by the settlor (e.g. ‘£100 to those who in the opinion of X are my friends’). Conceptual uncertainty is avoided because a conceptually uncertain class (‘friends’) is replaced by a conceptually certain class (‘friends as defined by X’).
In Re Tuck Eveleigh LJ held that the Chief Rabbi was of the latter type and so he was able to say the trust was valid.
* Is any party’s say-so sufficient to be used as the reference point through which conceptual certainty is obtained, or must the third party have so expertise or knowledge of the class set down by the settlor? Perhaps we the best approach is to ask whether we think it feasible for the stipulated third party to be able to resolve these difficulties. * Are all conceptually uncertain terms curable in this way? The third party may not be able to form an opinion on which of the different meanings of a term was intended by the settlor.
Some concepts are so ambiguous that, even if you were to limit the enquiry to one person’s conception of the term, there would still be uncertainty. E.g. Tall, attractive and intelligent are unlikely to be concrete conceptions. * One can argue that the distinction between cases in which the third party is called on to resolve conceptual uncertainty in a definition, and cases in which the third party’s opinion forms part of that definition so as to avoid uncertainty, is overly fine.
The courts have, at least on occasion, rescued a definition from conceptual uncertainty by interpreting it in such a way as to make its meaning clear (Re Gulbenkian’s Settlements, Lord Upjohn).
* Re Gulbenkian’s Settlements Facts: The objects of the power included all those with whom a named individual had been ‘residing’. Lord Reid argued that this was uncertain – ‘it is often difficult in a particular case to determine whether temporary sojourn amounts to ‘residence’. Decision: Held that such difficulties could be overcome and that, if necessary, the court could determine whether an individual matched the description
A similar exercise is evident in Re Baden’s Trusts (No.2) where the court held that ‘dependants’ was sufficiently certain. Sometimes the inherent imprecision in the terms themselves are underplayed by the court; i.e. Sachs LJ in Re Baden (No.2).
A trust may be void where it is so wide that the trust would be administratively unworkable. At the end of Lord Wilberforce’s judgment in McPhail v Doulton he implies that even where the ‘is or is not’ test is passed, the trust will fail if it is ‘administratively unworkable’.
What did Lord Wilberforce mean? Suggestions: * Settlor must make it clear not just who falls within the class of objects but also on the basis on which the trustees and/or court are to choose between them when it comes to distributing the trust fund. Without such info they don’t know who should get what. Problem:
Cases accept (although mostly implicitly) that there is no requirement that the settlor lay down the proper basis for distribution. He may leave it to the trustees to formulate their own criteria for selection. * It concerns the size of the class – if the class is too wide then it will fail. Gets superficial support from Lord Wilberforce’s example of a trust for all the residents of Greater London, although context suggests differently. Problem: Difficult to see why a large class of beneficiaries should in itself render a trust impossible to administer.
* Re Hay’s; Megarry VC indicated that, had he not already held the discretionary trust void on other grounds, he would have held it void as being administratively unworkable.
* R v. District Auditor No.3 ; Facts: The CC sought to create a discretionary trust whereby the trust fund of £400,000 was to be applied for certain purposes, ‘for the benefit of any or all or some of the inhabitants of the county of West Yorkshire. Decision: Lloyd LJ held, ‘A trust with as many as 2.5million potential beneficiaries is simply unworkable. This class size is far too large.’ Reasoning: Lloyd LJ believed that this case fell within Lord Wilberforce’s third category. Issues: No rationale is identified for the administrative unworkability requirement. Also, the trust should have failed on a separate ground – the rule prohibiting the creation of trusts for non-charitable purposes. ______________________________________________________________________________________________
Consequences of uncertainty/administrative unworkability
* If there is no certainty of intention then whoever has legal title to the relevant property keeps it beneficially. E.g. If S transfers property to T and it cannot be proved that S intended T to hold that property on trust, there is an outright transfer to T, who is then free to use the property as he likes. *
If there is uncertainty of subject matter or objects then, if the settlor sought to create the trust by declaring himself trustee, no trust arises and the settlor remains absolutely entitled to the property. If he sought to create the trust by transferring the property to another, T, for him to hold on trust, then uncertainty of subject matter or objects results in T holding that property on resulting trust for S. * Assuming the other two certainties are present, in any case where there is uncertainty of objects, the T cannot take beneficially, but will hold the trust property on a resulting trust for the settlor, or, where the trust arises under a will, for the persons entitled to the residue, or on